The Cyber Toolchain — issue 139, November 14, 2026
THE CYBER TOOLCHAIN
NO. 139
Tail
THE DAILY RELEASE FIREHOSE
PUBLISHED NOVEMBER 14, 2026 · EVERY DAY
EDITIONStail|grep|head|ls|diff|uniq
The daily firehose — everything the toolchain shipped today, already filtered.
We read every release from the 1325 tools on our watchlist at the source — GitHub and GitLab release notes, vendor release pages and changelogs, project blogs and feeds, vendor press releases, and the source code behind the tag. Bug-fix-only releases and non-product newsroom noise are dropped; what's left is summarized down to the new capability, how to try it, and any screenshots or videos the release itself published. Every entry links to the sources it was built from.
READ AS
Focus:What features shipped and what problems they solve, plus launches, packaging and where the market is moving.
Leaves out:Command-line flags, API-only deltas, usage examples, and releases that are only refinements.
Focus:New security controls, new APIs, and product improvements you can put to work, plus anything an upgrade breaks.
Leaves out:Launch and marketing news, governance, pricing, and documentation changes with no operational effect.
Focus:Dashboards, reporting, and anything you can quantify: metrics, coverage, risk scores, compliance impact, vendor moves.
Leaves out:Settings, flags, API and command-line detail, library bumps, and minor releases with no reporting or posture angle.
Focus:API, SDK and MCP changes, new libraries, breaking changes in what you import, and cyber tooling.
Leaves out:Governance, SOC operations, and vendor launches with no developer surface.
AMOUNT
DETAIL
ORDER
This issue came out before editions were ranked, so it shows the Standard edition in issue order.
What stands out across today's releases, grouped by what it lets you do. Every tool named links to its entry below.
Three unrelated things stand out. BBOT's new BlastDNS engine cuts subdomain scans by roughly 40% and turns CDNCDNA content delivery network, servers distributed across locations that cache and serve content near readers. It can reduce loading time and traffic to the original server./WAFWAFA reverse proxy that inspects HTTP/HTTPS traffic and blocks requests matching known attack patterns such as SQL injection or XSS, sitting between clients and a web application to filter malicious input before it reaches the server. port filtering on by default, so recon spends less time on dead ports. Finite State now flags KEVKEVCISA's Known Exploited Vulnerabilities catalog, a list of vulnerabilities with evidence of exploitation in real attacks. It helps teams prioritize fixes for problems attackers already use. membership on CVE feeds, so exploited bugs sort to the top without a second lookup. Findomain chains nuclei and ffuf straight onto discovered hosts.
identify
Finish an external attack-surface sweep in less wall-clock time, with fewer junk hosts
BBOT swaps massdns for BlastDNS for roughly 40% faster subdomain resolution and filters CDN/WAF ports by default, which kills the familiar problem of a scan reporting every edge IP as having 80/443 open. Findomain widens passive sourcing to 55 providers and moves default storage to SQLite, so large enumeration runs are queryable rather than a pile of text files.
Rank CVEs by what is actually being exploited before anyone opens a ticket
KEV membership now rides on the CVE update feed and findings sort by severity across five endpoints, so the triage step of cross-referencing CISA's catalog by hand disappears from the pipeline. Dependency-sourced BOMBOMA Bill of Materials listing every component, library, and dependency in a software project; cyber tools use it to track what is in a codebase and flag known-vulnerable components quickly. components also carry match evidence, which is what an engineer asks for first when disputing a transitive finding.
Scan container registries without parking a long-lived credential next to them
Registry Scanner authenticates to Azure Container Registry via managed or workload identityworkload identityAn identity assigned to software, such as a service or container, rather than a person. It lets the workload authenticate and receive permissions without sharing a user's credentials., removing the static registry secret that otherwise sits in config and never rotates. Coverage extends to ROSA with hosted control planes reading Amazon ECRAmazon ECRA managed container registry from AWS that stores, versions, and distributes Docker and OCI images; cyber tools use it to host and pull container images without running their own registry infrastructure..
See tunneled and encapsulated traffic that previously walked past the sensor
Maltrail 3.6 decapsulates ERSPANERSPANA Cisco-originated GRE-based tunneling protocol that encapsulates mirrored network traffic and ships it across Layer 3 to a remote capture point, letting cyber tools analyze traffic from distant segments without local tap hardware. and MPLSMPLSA packet-forwarding method that routes traffic using short numeric labels rather than full IP lookups, letting network operators engineer paths and stack tunnels with lower per-hop overhead. — the formats a span port or carrier link actually delivers — and adds heuristics for ICMP tunneling and IDN homographsIDN homographsCharacters from different scripts that look identical or nearly identical to Latin letters, used to craft domain names that visually impersonate legitimate ones; cyber tools flag them to detect phishing infrastructure., while a trail-store mmap cache drops cold reloads from 0.78s to 0.08s. eCapture now attaches process context to captured TLS traffic and writes pcapng by default, so you get HTTP reconstruction and the owning process without a conversion step.
Three unrelated things stand out today: JumpCloud turns access reviews into audit-ready campaigns, Runlayer adds a testable enforcement path for AI-agent policy, and Maltrail improves tunnel and homograph detection while making sensors easier to trust.
govern
Turn access reviews into evidence, not spreadsheet archaeology
Access CertificationAccess CertificationA periodic review process in which managers or system owners confirm or revoke which users should retain access to systems and data, reducing privilege creep over time. campaigns give identity teams a structured review flow with audit-ready output. That replaces ad hoc exports and email attestations with a repeatable control owners can show to auditors.
Prove AI-agent policy enforcement before relying on it
aiwatch enforce-test gives teams a verification path for policy enforcement, while OpenCode session monitoring extends Protect/Enforce coverage into another agent workflow. The practical win is catching gaps in AI-agent controls before developers assume the guardrail is active.
See tunneling and spoofing tricks in traffic you were already collecting
ERSPANERSPANA Cisco-originated GRE-based tunneling protocol that encapsulates mirrored network traffic and ships it across Layer 3 to a remote capture point, letting cyber tools analyze traffic from distant segments without local tap hardware. and MPLSMPLSA packet-forwarding method that routes traffic using short numeric labels rather than full IP lookups, letting network operators engineer paths and stack tunnels with lower per-hop overhead. decapsulation let sensors inspect encapsulated traffic instead of missing what was wrapped for transport, and new DNS CNAMECNAMEA DNS record type that maps one hostname to another as a canonical alias, letting cyber tools track infrastructure by following chains of names to their true origin., ICMP-tunnelingICMP-tunnelingA covert channel that encodes arbitrary data inside ICMP echo request/reply packets, exploiting firewalls that permit ping traffic to smuggle command-and-control or exfiltrated data past network controls. and IDNIDNInternationalized Domain Names — domain names containing non-ASCII characters encoded via Punycode so DNS can handle them; cyber tools flag them because they enable homograph phishing attacks using visually similar Unicode characters.-homograph heuristics add coverage for common evasion and deception patterns. The sensor health and alert-persistence layer also makes it easier to tell whether a quiet sensor is healthy or blind.
Prioritize vulnerable components with exploit context and evidence
Finite State Platform now flags KEVKEVCISA's Known Exploited Vulnerabilities catalog, a list of vulnerabilities with evidence of exploitation in real attacks. It helps teams prioritize fixes for problems attackers already use. membership in CVE update feeds and exposes match evidence for dependency-sourced BOMBOMA Bill of Materials listing every component, library, and dependency in a software project; cyber tools use it to track what is in a codebase and flag known-vulnerable components quickly. components, which helps teams separate known-exploited issues from theoretical backlog. Sysdig Secure adds credential-free Azure Container Registry scanning via managed or workload identityworkload identityAn identity assigned to software, such as a service or container, rather than a person. It lets the workload authenticate and receive permissions without sharing a user's credentials., removing long-lived scanner secrets from the registry-vulnerability workflow.
Hacktron's latest snapshot adds SLASLAA formal agreement between a service provider and customer that defines expected uptime, response times, and remedies for failures, giving cyber tools a contractual baseline for availability and support commitments. compliance tracking to its dashboard, no-findings report downloads for Whitebox pentests, and session-cookie login for pentest targets, alongside several integration and Slack UX fixes.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
└──▷ WHAT SHIPPED ·
7 FEATURES0 OF 7 FEATURESmost completely described firstwhat's the number?
Each feature carries 0–100 for how completely the vendor documented
it — not how big or important the work is. A major capability
described in eight words scores low, and that is the finding.
depth0–40what it does and how it works or what changed
specificity0–30names real surfaces — APIs, flags, formats, limits, numbers
actionability0–30enough to go use it — a named endpoint, flag, or config key tops this; a UI path is a starting point
Code and config rank first by construction: a documented endpoint,
flag or config key scores at the top of specificity and actionability,
so it sorts above a feature described only in prose. Under
60 the notes go thin — everything below the “thinner
coverage” line is thinner documentation, not smaller work.
Hover any meter for that feature's three sub-scores.
0100
SLA compliance ring in dashboardnew60how completely this was documenteddepth25/40specificity20/30actionability15/3060 / 100
The dashboard's PR review and Whitebox scan tabs now show an SLA compliance ring per severity, with MTTR and a Past SLA count linking to overdue findings.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
No-findings reports for Whitebox pentestsnew55how completely this was documenteddepth25/40specificity15/30actionability15/3055 / 100
Completed Whitebox pentests with no confirmed findings now offer Full and Executive report downloads, covering scope, methodology, zero counts, and a clear no-findings statement.
— Names report types and content but no exact download pathsnapshot-20261114
0300
Session-cookie auth for pentest login agentnew45how completely this was documenteddepth20/40specificity10/30actionability15/3045 / 100
Pentest auth instructions now accept pasted session cookies, so the login agent can sign in to a target without a password.
— Clear mechanism but no exact field or config name givensnapshot-20261114
0400
Correct author attribution in scan notificationsimproved45how completely this was documenteddepth20/40specificity15/30actionability10/3045 / 100
GitLab merge request and Bitbucket pull request scans now show the author as the creator in Slack finding messages and the scan list, instead of whoever connected the integration.
— Clear before/after, no configurable surface namedsnapshot-20261114
0500
Connection form UX for GitLab, Azure DevOps, GHEimproved45how completely this was documenteddepth20/40specificity15/30actionability10/3045 / 100
GitLab, Azure DevOps, and GitHub Enterprise Server connection forms now show collapsible Setup instructions and keep any connection error until you edit a field or retry.
— Describes UI change but no config key or field namedsnapshot-20261114
0600
Revoked connection detection for Bitbucketimproved40how completely this was documenteddepth15/40specificity10/30actionability15/3040 / 100
The Bitbucket integration now flags a revoked connection and offers to reconnect it from the integration sheet.
Xalgorix hardens its file tools against credential and secret leakage by expanding path-based deny-list protections, extending that deny-list to block writes as well as reads, and fixing ripgrep-based search to behave like grep -r.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
Xalgorix autonomous AI pentesting agents — real-time reconnaissance, vulnerability detection, and exploitation orchestration. Go + TypeScript.
Xalgorix hardens its file tools against credential and secret leakage by expanding path-based deny-list protections, extending that deny-list to block writes as well as reads, and fixing ripgrep-based search to behave like grep -r.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
└──▷ WHAT SHIPPED ·
3 FEATURES0 OF 3 FEATURESmost completely described firstwhat's the number?
Each feature carries 0–100 for how completely the vendor documented
it — not how big or important the work is. A major capability
described in eight words scores low, and that is the finding.
depth0–40what it does and how it works or what changed
specificity0–30names real surfaces — APIs, flags, formats, limits, numbers
actionability0–30enough to go use it — a named endpoint, flag, or config key tops this; a UI path is a starting point
Code and config rank first by construction: a documented endpoint,
flag or config key scores at the top of specificity and actionability,
so it sorts above a feature described only in prose. Under
60 the notes go thin — everything below the “thinner
coverage” line is thinner documentation, not smaller work.
Hover any meter for that feature's three sub-scores.
0100
Credential-file protections in file toolsnew85how completely this was documenteddepth30/40specificity30/30actionability25/3085 / 100
str_replace_editor, search_files, and list_files now refuse to touch xalgorix.env files, the provider key store and the auth profile store in Data_Dir, their temporary copies, and /proc/*/environ.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄ The read_deny field returned by /api/status now reflects these newly added deny-list entries so operators can confirm the protections are active.
Inspect the current deny-list to confirm new credential-file entries are active after upgrading.
— Names exact protected paths and a verifiable API field/commandv4.6.189
0200
Deny-list now blocks writes, not just readsbreaking80how completely this was documenteddepth35/40specificity25/30actionability20/3080 / 100
Paths listed in XALGORIX_READ_DENY_LIST now block file-tool writes as well as reads. This is a breaking change: any path previously allowed for writing via the write allow-list that also appears in XALGORIX_READ_DENY_LIST will now be refused on write.
— Clear before/after and exact env var, but no examplev4.6.189
thinner coverage below
0300
ripgrep search now matches grep -r behaviorimproved50how completely this was documenteddepth25/40specificity15/30actionability10/3050 / 100
search_files using ripgrep now searches the exact path it was given and includes hidden and ignored files, matching grep -r behavior.
— Describes behavior change but no command or flag givenv4.6.189
└──▷ BREAKING ON UPGRADE
!
Entries added through `XALGORIX_READ_DENY_LIST` now also block file-tool writes — any path previously allowed for writing via the write allow-list that also appears in `XALGORIX_READ_DENY_LIST` will be refused on write.Our best guess
OCInferno adds text-based OCI attack-path analysis, NSGNSGA Network Security Group is an Azure resource that filters inbound and outbound traffic to cloud resources using configurable allow/deny rules, giving cyber tools a policy layer to inspect or enforce network access control. security-rule enumeration, per-user DB credential enumeration, and identity propagation trust detection︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
└──▷ TRY IT
Find the shortest privilege-escalation path from every principal to any node with MANAGE access, using graph data already collected by OCInferno.
$ modules run process_og_attack_paths --graph-json /path/to/oracle_cloud_hound.json --to-manage--summary
Enumerate every simple escalation path to a specific resource kind (e.g. OCIGroup) and emit results as JSON for downstream processing.
$ modules run process_og_attack_paths --graph-json /path/to/oracle_cloud_hound.json --to-kind OCIGroup --all-paths--json
Analyze OCI privilege-escalation paths without BloodHound using the new process_og_attack_paths module, with flags --graph-json, --to-manage, --to-kind, --all-paths, --max-paths, --summary, and --json to control scope and output.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
Enumerate Network Security Group security rules as part of enum_core_network enumeration coverage.
Detect unrestricted identity propagation trusts — trusts with no client_claim_values set — via the new TABLE_IDD_IDENTITY_PROPAGATION_TRUSTS table and ipt_unrestricted_client_claim check.
Track customer secret keys per identity domain user via the new TABLE_IDD_USER_CUSTOMER_SECRET_KEYS table.
+1 moreshow less
Output folders now use friendly NAME_last4 naming instead of raw OCIDs.
Analyze OCI privilege-escalation paths without BloodHound using the new process_og_attack_paths module, with flags --graph-json, --to-manage, --to-kind, --all-paths, --max-paths, --summary, and --json to control scope and output.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
Detect unrestricted identity propagation trusts — trusts with no client_claim_values set — via the new TABLE_IDD_IDENTITY_PROPAGATION_TRUSTS table and ipt_unrestricted_client_claim check.
Track customer secret keys per identity domain user via the new TABLE_IDD_USER_CUSTOMER_SECRET_KEYS table.
Findomain is a fast subdomain enumeration tool that discovers subdomains of a target domain using certificate transparency logs and various APIs.
Findomain 11.0.0 adds built-in nuclei and ffuf scanning against discovered hosts, expands passive source coverage to 55 providers, and switches its default storage backend to SQLite while replacing the built-in port scanner with nmap — alongside SMTPSMTPThe application-layer protocol used to send email between servers and from clients to servers; cyber tools use it to relay alerts, phishing simulations, or exfiltrated data over standard mail infrastructure. email reports, webhook alerting for nuclei findings, and a wide set of breaking changes to flags, config formats and the library API.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
└──▷ WHAT SHIPPED ·
15 FEATURES4 OF 15 FEATURESmost completely described firstwhat's the number?
Each feature carries 0–100 for how completely the vendor documented
it — not how big or important the work is. A major capability
described in eight words scores low, and that is the finding.
depth0–40what it does and how it works or what changed
specificity0–30names real surfaces — APIs, flags, formats, limits, numbers
actionability0–30enough to go use it — a named endpoint, flag, or config key tops this; a UI path is a starting point
Code and config rank first by construction: a documented endpoint,
flag or config key scores at the top of specificity and actionability,
so it sorts above a feature described only in prose. Under
60 the notes go thin — everything below the “thinner
coverage” line is thinner documentation, not smaller work.
Hover any meter for that feature's three sub-scores.
0100
Nuclei vulnerability scanning integrationnew90how completely this was documenteddepth30/40specificity30/30actionability30/3090 / 100
Run nuclei templates against discovered live hosts using --nuclei, with --nuclei-templates, --nuclei-severity, --nuclei-tags, --nuclei-exclude-templates, --nuclei-rate-limit, and --nuclei-timeout controlling the scan, plus repeatable --nuclei-arg to pass arbitrary nuclei flags through.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
Nmap replaces built-in port scannerbreaking90how completely this was documenteddepth30/40specificity30/30actionability30/3090 / 100
Port scanning now shells out to nmap via --pscan, --nmap-full, --nmap-syn, --min-rate, --nmap-host-timeout, and passthrough --nmap-arg (or nmap_args in the config file); the previous built-in scanner and its --parallel-ip-ports-scan, --tcp-connect-threads, and --tcp-connect-timeout flags are removed.
— Names old and new flags with clear migration11.0.0
0302
SQLite as default storage backendbreaking90how completely this was documenteddepth30/40specificity30/30actionability30/3090 / 100
Results are now stored in SQLite by default, writing to ./findomain.db when no database flags are set; use --sqlite <file> or FINDOMAIN_SQLITE_PATH to choose a path. To keep using PostgreSQL, pass at least one of --postgres-user, --postgres-password, --postgres-host, --postgres-port, --postgres-database, or set postgres_connection in the config file or FINDOMAIN_POSTGRES_CONNECTION in the environment.
— Default change plus full opt-out path named11.0.0
0400
Time-budgeted passive source queriesnew88how completely this was documenteddepth30/40specificity28/30actionability30/3088 / 100
New --source-timeout (default 30s per request), --source-budget (default 300s total), and --archive-budget (default 20s for Wayback, CommonCrawl, UK Web Archive, and Arquivo) flags let operators bound how long Findomain waits on passive sources.
ffuf content discovery integrationnew86how completely this was documenteddepth28/40specificity28/30actionability30/3086 / 100
Run ffuf against live hosts with --ffuf, --ffuf-wordlist, --ffuf-threads, --ffuf-recursion, and --ffuf-recursion-depth, plus repeatable --ffuf-arg for arbitrary ffuf flags.
New reporting and alerting integrationsnew81how completely this was documenteddepth25/40specificity28/30actionability28/3081 / 100
Send scan reports over SMTP using --smtp-server, --smtp-port, --smtp-user, --smtp-password, and repeatable --email-to (including new subdomains plus nuclei and ffuf findings), and route nuclei findings to a dedicated Discord or Slack channel using smart_alerts_webhook in the config file or FINDOMAIN_SMART_ALERTS_WEBHOOK in the environment, capping the total time Findomain waits on a rate-limited webhook per message using --webhook-max-wait or webhook_max_wait.
— Flags named, content of reports described Both config keys and flag named11.0.0
0703
External subdomain import replaces built-in enumerationbreaking75how completely this was documenteddepth25/40specificity25/30actionability25/3075 / 100
--external-subdomains is removed; Findomain no longer runs amass or subfinder for you — use --import-subdomains to bring in their output files instead.
— Old and new flag named with clear migration11.0.0
0800
Passive source list expanded to 55 sourcesnew73how completely this was documenteddepth25/40specificity28/30actionability20/3073 / 100
Findomain now queries 55 passive sources (up from 14), including ahrefs, censys, shodan, fofa, intelx, leakix, netlas, and 35 others; any source can be excluded by identifier with --exclude-sources.
— Concrete count and named sources, exclude flag given11.0.0
0900
ARM64 Docker image and musl binarynew65how completely this was documenteddepth20/40specificity25/30actionability20/3065 / 100
The Docker image now builds for linux/arm64 in addition to linux/amd64, and a static findomain-aarch64-musl.zip asset is published for Alpine and other musl-based systems.
Windows release asset renamingbreaking65how completely this was documenteddepth20/40specificity25/30actionability20/3065 / 100
Windows release assets are renamed to findomain-windows.zip and findomain-windows-i686.zip; downloads from the old releases/latest/download/findomain-windows.exe.zip URL will now return 404.
Source identifier consolidation for crtsh and threatcrowdbreaking65how completely this was documenteddepth20/40specificity25/30actionability20/3065 / 100
crtsh is now a single source identifier covering both the database mirror and the JSON API, so --exclude-sources crtsh excludes both; the threatcrowd source is removed entirely.
`--external-subdomains` is removed; Findomain no longer runs amass or subfinder for you — use `--import-subdomains` to bring in their output files instead.Our best guess
SQLite is now the default database backend, writing to `./findomain.db`; to keep using PostgreSQL, pass at least one of `--postgres-user`, `--postgres-password`, `--postgres-host`, `--postgres-port`, `--postgres-database`, or set `postgres_connection` in the config file or `FINDOMAIN_POSTGRES_CONNECTION` in the environment.Our best guess
The built-in port scanner is replaced by nmap; `--parallel-ip-ports-scan`, `--tcp-connect-threads`, and `--tcp-connect-timeout` are removed and replaced by `--nmap-full`, `--nmap-syn`, `--min-rate`, `--nmap-host-timeout`, and `--nmap-arg`.Our best guess
Windows release assets are renamed to `findomain-windows.zip` and `findomain-windows-i686.zip`; downloads from `releases/latest/download/findomain-windows.exe.zip` will return 404.Our best guess
`crtsh` is now a single source identifier covering both the database mirror and the JSON API, so `--exclude-sources crtsh` excludes both.Our best guess
Fluxion is a wireless security auditing tool that creates rogue access points to capture WPA/WPA2 passwords via social engineering attacks.
CVE-2026-125399: Zero-Click, Unauthenticated OS Command Injection via eval() on Attacker-Controlled ESSID in Fluxion (CVSSCVSSCommon Vulnerability Scoring System, a method for describing a vulnerability's characteristics and severity. Its score is not the probability that attackers will exploit it. 8.8, affects <= 6.32)︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
GitGuardian SaaS is a secrets detection platform that scans Git repositories and developer workflows to find and remediate exposed credentials and sensitive data.
GitGuardian adds a secret revocation API, richer public incident fields, new audit log events for login lockout, and expanded perimeter bulk-selection controls.
Revoke a secret via the new POST /v1/secrets/{secret_id}/revoke endpoint, passing provider credentials in the credentials body field to override workspace defaults.
Filter public incidents and check runs by commit_sha, pull_request_number, or source_id in addition to existing filters; each check run now also returns its repository's full name and source ID.
Public incident responses now include company_related_verdict (values: related, unclear, unrelated) and agents_stage in list, detail, and CSV export, and both fields are filterable.
Filter the Teams page by Synced from provider when Team Sync is enabled, to distinguish synced teams from manually created ones.
Filter public incidents by Source owner — the GitHub organization or user, or Docker Hub namespace, that published the source.
+2 moreshow less
Lock out password logins for six minutes after ten failed attempts for the same email and IP address; the login endpoint responds with a Retry-After header and logs a password_login_locked_out entry in the audit log with the IP address and attempt count.
Select all sources matching current filters at the root level of an integration's perimeter to monitor or unmonitor them in one action.
SharePoint Online and OneDrive integrations no longer require the `Organization.Read.All` permission on your Entra application; installations without it are now named after their tenant ID rather than the organization name.Our best guess
Infisical is an open-source secret management platform that helps teams securely store, sync, and manage API keys, environment variables, and other sensitive configurations.
Infisical refined access request granularity for glob-based policies and loosened restrictions on self-hosted webhook alert destinations.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
└──▷ WHAT SHIPPED ·
2 FEATURES1 OF 2 FEATURESmost completely described firstwhat's the number?
Each feature carries 0–100 for how completely the vendor documented
it — not how big or important the work is. A major capability
described in eight words scores low, and that is the finding.
depth0–40what it does and how it works or what changed
specificity0–30names real surfaces — APIs, flags, formats, limits, numbers
actionability0–30enough to go use it — a named endpoint, flag, or config key tops this; a UI path is a starting point
Code and config rank first by construction: a documented endpoint,
flag or config key scores at the top of specificity and actionability,
so it sorts above a feature described only in prose. Under
60 the notes go thin — everything below the “thinner
coverage” line is thinner documentation, not smaller work.
Hover any meter for that feature's three sub-scores.
0101
Sub-path access requests within glob policiesimproved42how completely this was documenteddepth20/40specificity12/30actionability10/3042 / 100
Users can now request access to a narrower path inside a glob-based access policy, scoping the request to a specific secret path rather than the entire glob pattern.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
— Explains behaviour but no UI path or flag namedv0.167.10
0200
Self-hosted webhook alerts over HTTP and internal addressesimproved23how completely this was documenteddepth10/40specificity8/30actionability5/3023 / 100
Self-hosted Infisical instances can now send webhook alerts over plain HTTP and to internal addresses, which were previously restricted.
— Only mentioned in summary, no mechanism or config detail givenv0.167.10
TruffleHog finds, verifies, and analyzes leaked credentials across Git, chat, wikis, cloud storage, and more
TruffleHog Enterprise adds a Connections hub, per-source detail pages, a background CSV export API (POST /api/v3/secrets/exports, GET /api/v3/secrets/exports/{id}), and a low-memory Git scanning preview mode.
Adds a Connections page that groups Sources, Notifiers, Cloud Analyzers, and Scanners into one sortable, column-configurable view.
Adds a per-source detail page showing idle/scanning status, next scheduled scan, configuration, scanner details, triage status, and recently discovered live secrets.
Export secrets to CSV at scale via POST /api/v3/secrets/exports and poll GET /api/v3/secrets/exports/{id} for status; filters apply via query string.
Adds a low-memory Git scanning mode for scanners on memory-limited hosts (private preview).
The Integrations page is renamed to Connections across navigation and the product; existing links to the old Scanners page under Settings redirect to the new Scanners tab under Connections.Our best guess
GitHub Copilot gained GitHub Enterprise Server account support and a new Gemini 4 Hydrogen model option, while the REST APIREST APIA web API design style where clients interact with resources over HTTP using standard methods like GET, POST, and DELETE, giving cyber tools a predictable, language-agnostic way to integrate with external services. added repository-level controls for AI Scan.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
GitHub is a web-based platform for version control and collaborative software development using Git repositories.
GitHub Copilot gained GitHub Enterprise Server account support and a new Gemini 4 Hydrogen model option, while the REST APIREST APIA web API design style where clients interact with resources over HTTP using standard methods like GET, POST, and DELETE, giving cyber tools a predictable, language-agnostic way to integrate with external services. added repository-level controls for AI Scan.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
└──▷ WHAT SHIPPED ·
3 FEATURES0 OF 3 FEATURESmost completely described firstwhat's the number?
Each feature carries 0–100 for how completely the vendor documented
it — not how big or important the work is. A major capability
described in eight words scores low, and that is the finding.
depth0–40what it does and how it works or what changed
specificity0–30names real surfaces — APIs, flags, formats, limits, numbers
actionability0–30enough to go use it — a named endpoint, flag, or config key tops this; a UI path is a starting point
Code and config rank first by construction: a documented endpoint,
flag or config key scores at the top of specificity and actionability,
so it sorts above a feature described only in prose. Under
60 the notes go thin — everything below the “thinner
coverage” line is thinner documentation, not smaller work.
Hover any meter for that feature's three sub-scores.
0100
AI Scan enablement via REST APInew95how completely this was documenteddepth35/40specificity30/30actionability30/3095 / 100
New repository_scan property can be set to enabled or disabled on the AI Scan enablement endpoints, alongside pr_scan, in the same request via PATCH to /orgs/{org}/code-scanning/ai-scan or /repos/{owner}/{repo}/code-scanning/ai-scan; if either change is blocked a 422 error is returned and neither setting changes.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄ Organization-level repository_scan cascades to repositories: an org-level disable prevents repos from re-enabling it, an org-level enable lets repos inherit and individually opt out, and enterprise policies that disallow AI Scan cause enable requests to return 422.
Enable AI Scan for repositories across an entire organization while also enabling PR scan in one API call.
Gemini 4 Hydrogen model in Copilotnew75how completely this was documenteddepth28/40specificity25/30actionability22/3075 / 100
GitHub Copilot now supports Gemini 4 Hydrogen, Google's latest Flash model, selectable from the model picker in Visual Studio Code, Visual Studio, Copilot CLI, GitHub Copilot coding agent, the Copilot app, github.com, GitHub Mobile (iOS and Android), JetBrains IDEs, Xcode, Eclipse, and the Copilot SDK. Copilot Enterprise and Business administrators can manage access via the model policy in Copilot settings; under default model enablement, the model is automatically enabled unless the global default is off or the model is explicitly disabled.
Select Gemini 4 Hydrogen for a Copilot CLI session to run fast, multistep agentic coding tasks from the terminal.
📍In the Copilot CLI model picker, select Gemini 4 Hydrogen before issuing your request.
Restrict Gemini 4 Hydrogen access org-wide before its gradual rollout reaches all users.
📍In Copilot settings (Enterprise or Business admin), go to the model policy and explicitly disable Gemini 4 Hydrogen under Default model enablement.
GitHub Enterprise Server accounts in Copilot appnew72how completely this was documenteddepth25/40specificity22/30actionability25/3072 / 100
The GitHub Copilot app now generally supports connecting a GitHub Enterprise Server account to work with remote GHES repositories, issues, and pull requests, available in app version 1.2.0 and later. Add it via Settings > Accounts > Add account, then choose GitHub Enterprise Server, entering the GHES instance URL and credentials.
Add your GHES account to the Copilot app to access on-premises repositories, issues, and pull requests alongside your GitHub.com work.
📍1. Sign in with a GitHub.com or GHE.com account.
2. Go to Settings › Accounts › Add account.
3. Choose GitHub Enterprise Server and enter your GHES instance URL and credentials.
Syzkaller is an unsupervised coverage-guided kernel fuzzer developed by Google that automatically finds bugs in OS kernels by generating and executing random syscall sequences.
syzkaller adds cross-compilation and image-creation support for FreeBSD and OpenBSD on Linux, plus a /healthz endpoint and kernel_builder_image option for syz-ci︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
└──▷ TRY IT
Build an OpenBSD/amd64 target image, sysroot, and builder image on a Linux host for cross-compiled fuzzing.
Check whether a syz-hub instance is alive in a health-monitoring or deployment pipeline.
$ curl http://<syz-hub-host>:<port>/healthz
Configure a custom kernel builder container image in syz-ci with the new kernel_builder_image option.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
Add a /healthz endpoint to syz-hub for liveness checks and graceful shutdown support.
Cross-compile and build FreeBSD kernels on Linux hosts, using the new tools/create-freebsd-image.sh script to produce target disk images.
Cross-compile and build OpenBSD/amd64 kernels on Linux hosts, using the new tools/create-openbsd-image.sh script to produce target disk images, sysroots, and builder images.
syz-hub now validates manager names on registration.
Configure a custom kernel builder container image in syz-ci with the new kernel_builder_image option.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
Add a /healthz endpoint to syz-hub for liveness checks and graceful shutdown support.
Cross-compile and build FreeBSD kernels on Linux hosts, using the new tools/create-freebsd-image.sh script to produce target disk images.
Cross-compile and build OpenBSD/amd64 kernels on Linux hosts, using the new tools/create-openbsd-image.sh script to produce target disk images, sysroots, and builder images.
Snyk Code is a static application security testing (SAST) tool that scans source code in real time to identify and fix security vulnerabilities.
Snyk Code now reads project-own configuration to resolve imports in Go, TypeScript, and JavaScript, following data across packages and modules for deeper analysis.
Checkmarx agentic application security software platform Combines Hybrid scanning, AI-powered agents, and unified risk intelligence across every attack surface
Checkmarx adds an OpenAI Daybreak Defense Factory integration, piping SASTSASTStatic application security testing, analyzing source code or compiled code for vulnerabilities without running the application. It helps developers find problems before deployment. findings via the Codex Security CLI into AI-powered investigation and remediation workflows.
Import Checkmarx SAST findings into OpenAI Daybreak Defense Factory by running the Codex Security CLI with --provider checkmarx and --dry-run to preview, then without --dry-run to import.
New integrationSupported scan result formats for the Daybreak Defense Factory integration are Checkmarx One JSON and CxSAST XML reports, covering both Checkmarx One and self-managed CxSAST deployments.
SonarQube is an enforce consistent quality gates across every pull request and pipeline, from IDE to CI/CD, with automated code review for bugs and security issues.
SonarQube added an integration with OpenAI's Daybreak Defense Factory, letting Codex Security import SonarQube findings and validate fixes against the Sonar quality gate.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
└──▷ WHAT SHIPPED ·
1 FEATURE0 OF 1 FEATUREmost completely described firstwhat's the number?
Each feature carries 0–100 for how completely the vendor documented
it — not how big or important the work is. A major capability
described in eight words scores low, and that is the finding.
depth0–40what it does and how it works or what changed
specificity0–30names real surfaces — APIs, flags, formats, limits, numbers
actionability0–30enough to go use it — a named endpoint, flag, or config key tops this; a UI path is a starting point
Code and config rank first by construction: a documented endpoint,
flag or config key scores at the top of specificity and actionability,
so it sorts above a feature described only in prose. Under
60 the notes go thin — everything below the “thinner
coverage” line is thinner documentation, not smaller work.
Hover any meter for that feature's three sub-scores.
0100
Codex Security integration via Daybreak Defense Factorynew50how completely this was documenteddepth20/40specificity20/30actionability10/3050 / 100
SonarQube now works within OpenAI Daybreak Defense Factory through a SonarQube importer in Codex Security, a SonarQube plugin for Codex, and the SonarQube CLI, enabling security findings to flow into Codex Security in real time for validation and remediation against the Sonar quality gate.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
Semgrep shipped a beta MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface. server that lets AI coding agents triage findings, manage scans, and configure policy directly against the Semgrep AppSecAppSecApplication security: the practice of finding and fixing vulnerabilities in software during design, development, and testing. Cyber tools build on AppSec workflows to shift vulnerability detection earlier in the development lifecycle. Platform, alongside a wave of Guardian updates adding Copilot support, org-wide policy enforcement, AI-powered vulnerability detection, lockfile protection against supply chain attacks, and a staged-rollout Observe Mode.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
Get started with Semgrep to help you catch, flag, and fix real vulnerabilities before they ship.
Semgrep shipped a beta MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface. server that lets AI coding agents triage findings, manage scans, and configure policy directly against the Semgrep AppSecAppSecApplication security: the practice of finding and fixing vulnerabilities in software during design, development, and testing. Cyber tools build on AppSec workflows to shift vulnerability detection earlier in the development lifecycle. Platform, alongside a wave of Guardian updates adding Copilot support, org-wide policy enforcement, AI-powered vulnerability detection, lockfile protection against supply chain attacks, and a staged-rollout Observe Mode.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
└──▷ WHAT SHIPPED ·
5 FEATURES5 OF 5 FEATURESmost completely described firstwhat's the number?
Each feature carries 0–100 for how completely the vendor documented
it — not how big or important the work is. A major capability
described in eight words scores low, and that is the finding.
depth0–40what it does and how it works or what changed
specificity0–30names real surfaces — APIs, flags, formats, limits, numbers
actionability0–30enough to go use it — a named endpoint, flag, or config key tops this; a UI path is a starting point
Code and config rank first by construction: a documented endpoint,
flag or config key scores at the top of specificity and actionability,
so it sorts above a feature described only in prose. Under
60 the notes go thin — everything below the “thinner
coverage” line is thinner documentation, not smaller work.
Hover any meter for that feature's three sub-scores.
0101
MCP server for AI coding agentsNew integrationnew94how completely this was documenteddepth38/40specificity28/30actionability28/3094 / 100
Connect Claude Code, Codex, Cursor, VS Code, or ChatGPT as an MCP client to Semgrep AppSec Platform.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄ Agents can search and filter findings from Semgrep Code, Supply Chain, Secrets, and Agentic Workflows, triage and reopen findings in bulk, and read code snippets and finding history; start managed scans, find failed scans, read scan logs and errors, and retry scans; look up advisories, list dependencies, and find every repository affected by an advisory; dry-run and apply policy changes, manage exceptions, and configure Semgrep Guardian rules; pull reports by severity, product, or rule including median open age, findings activity, and guardrails adoption; trigger and schedule Agentic Workflows (admins only), read job status and task logs, and track credit usage; and create and link Jira and Linear tickets, build automations, view Slack notification settings, and check and sync SCMSCMSource Code Management: version-control systems (Git, SVN, etc.) that track changes to code over time, giving cyber tools reproducible builds, audit trails, and collaborative development workflows. connections. Authenticate via OAuth (browser sign-in scoped to an organization) or an API token — a Web API token allows all read and write operations, a Read-only token allows reads only, and an Agent (CI) token covers read and write operations on scans, while Member tokens are not supported. A discover tool finds the right MCP tool for a task and a semgrep_account tool returns the connected organization. Available to organizations on Teams or Enterprise plans (trials included) using any MCP client that supports remote servers over HTTP.
Connect Cursor (or any MCP client that supports remote HTTP servers) to the Semgrep MCP server so your agent can triage findings and manage policies without leaving the editor.
📍In Cursor, open Settings › MCP and add the Semgrep MCP server URL. Choose OAuth to sign in through your browser and scope access to your organization, or supply a Web API token for full read and write operations. Once connected, your agent can call the discover tool to list available capabilities or semgrep_account to confirm which organization it is connected to.
— Extensive named tools, token types and client list with a connection walkthrough.launch-20261112-f2d6b2d0
0202
Guardian installation across coding agentsnew83how completely this was documenteddepth25/40specificity28/30actionability30/3083 / 100
Guardian can now be installed in Claude Code via claude plugin install semgrep@claude-plugins-official, in Codex via codex plugin marketplace add https://github.com/semgrep/guardian --sparse .claude-plugin --sparse plugin followed by codex plugin add semgrep@semgrep-marketplace, and in Copilot via copilot plugin marketplace add https://github.com/semgrep/guardian followed by copilot plugin install guardian-copilot@semgrep-marketplace, extending Guardian beyond the previously supported Codex and Claude Code agents to Microsoft Copilot.
Bootstrap Guardian in Claude Code so every future session enforces your org's Semgrep policies automatically.
$ claude plugin install semgrep@claude-plugins-official
Add Guardian to Codex so the agent can detect, fix, and report Semgrep findings without leaving the agent environment.
Org-wide policy enforcement in Guardiannew72how completely this was documenteddepth28/40specificity22/30actionability22/3072 / 100
Guardian's org-wide policy system lets security teams define custom YAML rules (such as a no-legacy-auth rule) in a local rules file or a repo-wide file, automatically applying them to every developer's agent session without manual distribution.
Define a custom org rule that blocks a deprecated authentication library, then commit it to your local rules file so Guardian enforces it in every agent session.
yaml
rules:
- id: no-legacy-auth
languages: [python]
severity: ERROR
message: Use auth.v2.verify() — legacy_auth is deprecated.
pattern: legacy_auth(...)
Lockfile protection against supply chain attacksnew62how completely this was documenteddepth30/40specificity24/30actionability8/3062 / 100
Guardian blocks agent-initiated changes to lockfiles that would introduce malware or a reachable vulnerability, preventing unsafe dependency changes before they reach the build system, and pushes advisories to every developer to check whether their installed version is still vulnerable when a critical supply chain incident (such as PHP2Shell or Node2Shell) is disclosed.
— Named incidents and mechanism, but it's automatic with no reader action.launch-20261112-c889153f
thinner coverage below
0505
AI-powered vulnerability detection in GuardianNew detectionnew51how completely this was documenteddepth25/40specificity18/30actionability8/3051 / 100
Guardian's AI-powered detection reviews code at generation time and reasons about complex vulnerabilities including insecure direct object references (IDORIDORInsecure Direct Object Reference, a web vulnerability where an application exposes internal identifiers (IDs, filenames) without access checks, letting an attacker substitute another user's ID to reach their data.) and broken object-level authorization (BOLABOLABroken Object Level Authorization, an API vulnerability class where a server fails to verify that the requesting user owns or may access the object ID they supply, letting attackers read or modify other users' data.), as well as code written with malicious intent.
Semgrep adds Mandoline, an agent-oriented static-analysis engine with CLI, MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface., and gRPCgRPCAn open-source remote procedure call framework from Google that uses HTTP/2 and Protocol Buffers to let services call each other's functions across a network, giving tools low-latency, strongly typed inter-process communication. interfaces, callgraph/dataflow/taint primitives, and UniFFI bindings for six languages.
Access Mandoline via CLI, MCP, or gRPC interfaces, with callgraph output available in text, JSON, GRAPHVIZ, and protobuf formats.
Semgrep adds Mandoline, a decomposed static-analysis engine built for agentic workflows, exposing callgraph, dataflow, taint paths, API endpoint detection, and dependency reachability as composable primitives.
Mandoline adds native library bindings for six languages via UniFFI, plus WASMWASMA binary instruction format for a stack-based virtual machine, maintained by the W3C, that lets code compiled from any language run at near-native speed inside a sandboxed environment across browsers and servers. support, enabling use in agent containers, edge compute, and the browser or IDE.
Semgrep customers now have access to Mandoline inside Semgrep Agentic Workflows and Semgrep's Context Layer.
Semgrep adds Mandoline (a millisecond-speed program analysis engine), Context Layer (private beta, persistent codebase context), Agentic Workflows, and Semgrep MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface. for agent-native AppSecAppSecApplication security: the practice of finding and fixing vulnerabilities in software during design, development, and testing. Cyber tools build on AppSec workflows to shift vulnerability detection earlier in the development lifecycle..
Semgrep adds Mandoline, a next-generation program analysis engine that answers questions about code flows and reachable paths in milliseconds, scoring 100% on the OWASPOWASPA nonprofit foundation that publishes open standards, guides, and ranked vulnerability lists—most famously the OWASP Top 10—giving cyber tools a vendor-neutral benchmark to test or report against. Benchmark for Java and 98.28% on Python in a single run with no AI token usage.
Semgrep adds Context Layer, now in private beta, giving agentic workflows persistent, dynamic understanding of a codebase and organization across runs, enabling agents to surface 2x–7x more real vulnerabilities compared to agents without it.
Semgrep adds Agentic Workflows, powered by Mandoline and Context Layer, to automate complex detection, triage, and fix tasks following guided paths set by Semgrep's security experts.
Access the full Semgrep platform from any agent by pointing it at the Semgrep MCP server, available in Codex, Claude Code, Cursor, and CLI.
Semgrep adds deployment options including bring-your-own-key (BYOKBYOKBring Your Own Key — a cloud deployment model where the customer supplies and controls the encryption keys rather than the provider, so sensitive data stays under the customer's cryptographic authority.) models, Agentic Workflows in CI, Semgrep Dedicated single-tenant hosting, and an EU multi-tenant region.
Cloudsmith is a fully-managed artifact management, software supply chain security, and software distribution platform.
Cloudsmith's CLI picked up a new AI skills registry command surface, credential helpers for Maven/Gradle/sbt, and quality-of-life fixes for deprecated option warnings and SAMLSAMLA standard for passing signed identity assertions between a sign-in provider and an application. It lets a user sign in through their organization's identity provider. workspace persistence.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
└──▷ WHAT SHIPPED ·
4 FEATURES0 OF 4 FEATURESmost completely described firstwhat's the number?
Each feature carries 0–100 for how completely the vendor documented
it — not how big or important the work is. A major capability
described in eight words scores low, and that is the finding.
depth0–40what it does and how it works or what changed
specificity0–30names real surfaces — APIs, flags, formats, limits, numbers
actionability0–30enough to go use it — a named endpoint, flag, or config key tops this; a UI path is a starting point
Code and config rank first by construction: a documented endpoint,
flag or config key scores at the top of specificity and actionability,
so it sorts above a feature described only in prose. Under
60 the notes go thin — everything below the “thinner
coverage” line is thinner documentation, not smaller work.
Hover any meter for that feature's three sub-scores.
0100
Maven, Gradle, and sbt credential helpersnew50how completely this was documenteddepth20/40specificity15/30actionability15/3050 / 100
The Cloudsmith CLI now sets up Maven, Gradle, and sbt to fetch credentials directly from the CLI, so build files no longer need a stored API key.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
StepSecurity secures GitHub Actions CI/CD pipelines by hardening workflows, detecting anomalies, and preventing supply chain attacks through automated security controls.
StepSecurity adds a Critical-severity dashboard control that flags weak fork PR approval policies at both org and repo level, with hourly refresh.
Each finding distinguishes whether the weak policy comes from the organization's default or a repository-level override, so repositories that deviate from the org setting are caught separately.
New detectionFindings update on the next hourly scan after the setting changes; to remediate, set the policy to Require approval for all external contributors under Settings > Actions > General in GitHub.
Burp Suite is a we enable the world to secure the web. Burp Suite, the Web Security Academy, and world-leading research from PortSwigger â the AppSec communityâs home.
Burp Suite 2026.11 adds a new Academy tab for launching Web Security Academy labs directly from the tool, alongside easier project switching and restarting.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
└──▷ WHAT SHIPPED ·
2 FEATURES0 OF 2 FEATURESmost completely described firstwhat's the number?
Each feature carries 0–100 for how completely the vendor documented
it — not how big or important the work is. A major capability
described in eight words scores low, and that is the finding.
depth0–40what it does and how it works or what changed
specificity0–30names real surfaces — APIs, flags, formats, limits, numbers
actionability0–30enough to go use it — a named endpoint, flag, or config key tops this; a UI path is a starting point
Code and config rank first by construction: a documented endpoint,
flag or config key scores at the top of specificity and actionability,
so it sorts above a feature described only in prose. Under
60 the notes go thin — everything below the “thinner
coverage” line is thinner documentation, not smaller work.
Hover any meter for that feature's three sub-scores.
0100
Academy tab for launching labs in-toolnew50how completely this was documenteddepth20/40specificity15/30actionability15/3050 / 100
A new Academy tab lets users access and launch Web Security Academy labs directly inside Burp Suite without switching to a browser.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄ Users click the Academy tab in the top navigation bar to browse and launch labs.
Access Web Security Academy labs directly inside Burp Suite without switching to a browser.
📍In Burp Suite, click the Academy tab in the top navigation bar to browse and launch Web Security Academy labs.
Official documentation for Invicti AppSec Platform, Invicti Platform, and Invicti Enterprise — web application security testing, DAST, API security, and vulnerability management.
Invicti adds target type filtering on the Scans page, configurable default Accept headers per target, removed-operation tracking in the API catalog, and API spec discovery in Source Scan.
Filter and sort scans by target type (Web or API) using the new target type filter and column on the Scans page.
Set the default Accept header the scanner sends on web and API requests using the Default Web Accept Header and Default API Accept Header settings in your target's advanced settings.
Source Scan now finds existing API specs in your source code and uses them to build a more complete picture of the API your application exposes.
API catalog now marks operations that no longer appear in a re-imported API definition as Removed, with a status filter to show or hide them.
Zero Configuration discovery is now enabled by default, requiring no extra setup.
+2 moreshow less
Email notifications for new and updated APIs are now off by default, requiring opt-in via profile notification settings.
Updated integrationTest the connection when saving an API gateway integration to catch connection problems immediately.
Fastly is a cloud edge platform providing CDN, DDoS protection, load balancing, and real-time content delivery to accelerate and secure web applications.
Fastly's JavaScript SDK 4.3.0 adds new cryptographic key import formats, a mid-request handoff capability for compute requests, and stabilizes previously experimental APIs into standard modules.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
└──▷ WHAT SHIPPED ·
3 FEATURES0 OF 3 FEATURESmost completely described firstwhat's the number?
Each feature carries 0–100 for how completely the vendor documented
it — not how big or important the work is. A major capability
described in eight words scores low, and that is the finding.
depth0–40what it does and how it works or what changed
specificity0–30names real surfaces — APIs, flags, formats, limits, numbers
actionability0–30enough to go use it — a named endpoint, flag, or config key tops this; a UI path is a starting point
Code and config rank first by construction: a documented endpoint,
flag or config key scores at the top of specificity and actionability,
so it sorts above a feature described only in prose. Under
60 the notes go thin — everything below the “thinner
coverage” line is thinner documentation, not smaller work.
Hover any meter for that feature's three sub-scores.
0100
SPKI and PKCS8 key import formatsnew65how completely this was documenteddepth20/40specificity25/30actionability20/3065 / 100
Fastly's JavaScript SDK now supports spki and pkcs8 formats in SubtleCrypto.importKey for importing public and private keys.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
Fastly now lets you apply best-practice configurations when creating a workspace in the control panel.
Apply a set of best-practice configurations at workspace creation time, including immediate blocking of requests tagged with attack signals and rules covering malformed payloads and known malicious IP addresses.
Request Tracker 6.0.4 reworked how tickets handle attachments and links, adding direct uploads with pinning and bulk download, plus a redesigned Links section with type-aware autocomplete and inline field editing.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
Request Tracker / RTIR (Best Practical) is an open source ticketing system from Best Practical. Self-host RT or let us run it in the cloud.
Request Tracker 6.0.4 reworked how tickets handle attachments and links, adding direct uploads with pinning and bulk download, plus a redesigned Links section with type-aware autocomplete and inline field editing.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
└──▷ WHAT SHIPPED ·
2 FEATURES0 OF 2 FEATURESmost completely described firstwhat's the number?
Each feature carries 0–100 for how completely the vendor documented
it — not how big or important the work is. A major capability
described in eight words scores low, and that is the finding.
depth0–40what it does and how it works or what changed
specificity0–30names real surfaces — APIs, flags, formats, limits, numbers
actionability0–30enough to go use it — a named endpoint, flag, or config key tops this; a UI path is a starting point
Code and config rank first by construction: a documented endpoint,
flag or config key scores at the top of specificity and actionability,
so it sorts above a feature described only in prose. Under
60 the notes go thin — everything below the “thinner
coverage” line is thinner documentation, not smaller work.
Hover any meter for that feature's three sub-scores.
0100
Direct attachment uploads with pinning and bulk downloadimproved70how completely this was documenteddepth30/40specificity20/30actionability20/3070 / 100
Attachments can now be uploaded directly from the ticket display page without starting a reply or comment, and can be renamed, pinned, or deleted from that same page.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄ Pinning keeps key files at the top of the attachment list, and the list now supports multiple sort options, filtering, and bulk actions including bulk download.
Pin the current contract to the top of a ticket's attachment list so the whole team can find it without scrolling.
📍On the ticket display page, go to the Attachments section, upload or locate the contract file, then select Pin from its action menu to keep it at the top of the list.
Bulk-download all attachments from a ticket to archive a closed project's files in one action.
📍On the ticket display page, go to the Attachments section, select all files using the bulk-select control, then choose Bulk download from the bulk actions menu.
Redesigned Links section with autocomplete and inline editimproved70how completely this was documenteddepth30/40specificity20/30actionability20/3070 / 100
The Links section now shows a dropdown of all linkable RT object types — assets, articles, and other objects — plus an option to add a plain web link. After choosing a link type, the input field becomes an autocomplete search box, and users can choose which fields appear for linked tickets and edit those fields inline without leaving the current ticket.
Link an asset to a ticket and update one of its fields inline without opening the asset's own page.
📍On the ticket display page, go to the Links section, open the link-type dropdown and select the asset object type, search for the asset by name in the autocomplete box, add the link, then click the desired field on the linked asset row to edit it inline.
Accelerate your digital forensics and incident response investigations with Belkasoft
Belkasoft X 2.13 adds MFTMFTThe Master File Table, a core NTFS structure maintained by Windows that records metadata for every file and directory on a volume, giving cyber tools a fast path to forensic artifacts without full disk traversal.-based deleted file recovery alongside a wave of BelkaGPT upgrades — document summarization, email/SMSSMSShort Message Service, the protocol built into cellular networks for sending text messages between phones; cyber tools use it for alerting, two-factor authentication delivery, or phishing simulation over mobile channels. topic detection, faster Hub- and CPUCPUThe main general-purpose processor in a computer. It runs application instructions and coordinates work with memory and other hardware.-based answer generation with chain-of-thoughtchain-of-thoughtIntermediate reasoning a model writes out before its final answer. Reasoning models generate it automatically and it is often billed as extra output tokens. display — plus browser-based Google Cloud sign-in, dump.bin support for MTK acquisition, and new mobile app artifact coverage.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
└──▷ WHAT SHIPPED ·
9 FEATURES0 OF 9 FEATURESmost completely described firstwhat's the number?
Each feature carries 0–100 for how completely the vendor documented
it — not how big or important the work is. A major capability
described in eight words scores low, and that is the finding.
depth0–40what it does and how it works or what changed
specificity0–30names real surfaces — APIs, flags, formats, limits, numbers
actionability0–30enough to go use it — a named endpoint, flag, or config key tops this; a UI path is a starting point
Code and config rank first by construction: a documented endpoint,
flag or config key scores at the top of specificity and actionability,
so it sorts above a feature described only in prose. Under
60 the notes go thin — everything below the “thinner
coverage” line is thinner documentation, not smaller work.
Hover any meter for that feature's three sub-scores.
0100
BelkaGPT document summarizationnew93how completely this was documenteddepth35/40specificity28/30actionability30/3093 / 100
BelkaGPT now summarizes documents across word processor files, PDFs, spreadsheets, slide decks, and plain-text logs via Summarize with BelkaGPT in the Artifacts window's Documents profile, with Quick summary covering the first 500 words and Full summary covering up to 25,000 words, in a selectable language, with offload support via BelkaGPT Hub.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
Quickly assess a large batch of documents for investigative relevance without reading each one in full.
📍Select all items under the Documents profile in the Artifacts window, right-click and choose Summarize with BelkaGPT, select Full summary (up to 25,000 words) or Quick summary (first 500 words), pick the output language, and optionally route processing to BelkaGPT Hub if the local workstation lacks a compatible GPU.
BelkaGPT topic detection for emails and SMSimproved86how completely this was documenteddepth30/40specificity28/30actionability28/3086 / 100
BelkaGPT topic detection, previously available only for chats, now also covers emails and SMS messages, using the same predefined detectors (money, drugs, operational coordination, account access) and custom detectors already available for chats, run from the BelkaGPT Topic Detection panel.
Surface relevant email and SMS threads matching investigative topics — such as drug references or operational coordination — across all communication types at once.
📍In the BelkaGPT panel, open Topic Detection, select Emails and SMS as sources alongside any existing chat sources, choose a predefined detector (for example, Drugs or Operational coordination) or add a custom detector describing case-specific coded language, then run detection to collect matching messages.
MFT-based deleted file analysisnew85how completely this was documenteddepth30/40specificity25/30actionability30/3085 / 100
Belkasoft X 2.13 adds MFT-based deleted file analysis for NTFSNTFSThe default file system format for Windows, developed by Microsoft, storing files with rich metadata and access controls; cyber tools target it to parse artifacts like timestamps, alternate data streams, and the MFT. volumes, surfaced in the File System window under a new Deleted items node with Orphans (unknown parent path) and Path restored (reconstructed full path) subnodes, filterable by name, size, or timestamp, with recovery via Save.
Locate and recover deleted files from an NTFS volume without leaving the case — useful when a suspect wiped files but data may still be on disk.
📍In the File System window, expand the target NTFS volume, then open the Deleted items node. Browse Orphans for files whose parent path is unknown or Path restored for files whose full original path was reconstructed. Filter by name, size, or timestamp columns to narrow results, then right-click a file and choose Save to recover it.
BelkaGPT answer generation: Hub offload, fast CPU mode, chain of thoughtimproved70how completely this was documenteddepth30/40specificity25/30actionability15/3070 / 100
BelkaGPT answers can now be generated via BelkaGPT Hub to offload Q&A processing from the local forensic workstation, and a new fast answer mode on CPU reduces answer time to about 1 to 3 minutes (down from 12 to 15 minutes) on machines without a compatible GPUGPUA graphics processing unit, a processor that performs many calculations in parallel. AI workloads use GPUs to speed up model training and inference. or Hub. BelkaGPT also now shows chain of thought during answer generation, displaying each reasoning step and the time it took.
MTK acquisition dump.bin support and decryption utilitynew70how completely this was documenteddepth25/40specificity25/30actionability20/3070 / 100
MTK acquisition now supports dump.bin files added directly to a case as a mobile image, as an alternative to .belkaml images. A new utility, available on request from support@belkasoft.com, creates decryption material for MTKClient dumps so they can be decrypted in Belkasoft X.
New Android and iOS artifact supportnew60how completely this was documenteddepth20/40specificity30/30actionability10/3060 / 100
Belkasoft X adds Android artifact support for Claude AI v1.260625.20, Copilot v30.0.440623001, ChatGPT v5.2, Google Files v1.2263, Google Messages, Google Photos, Samsung Gallery Trash v14.5.12, and Samsung My Files v14.5, plus iOS artifact support for Claude AI v1.260618.1 and Copilot v30.0.440608001.
BelkaGPT classification results in Evidence Reader exportsimproved55how completely this was documenteddepth20/40specificity20/30actionability15/3055 / 100
BelkaGPT classification results — topic detection, SQLite database classification, and picture classification — are now included in exports from the Artifacts window to Belkasoft Evidence Reader.
CAPEv2 is an open-source malware sandbox that automatically executes and analyzes suspicious files to extract configurations, payloads, and behavioral indicators.
CAPEv2 adds tshark-based network processing with TLS session key parsing from tlsdump.log and NSS keylogNSS keylogA file that records TLS session keys in NSS (Network Security Services) format, enabling tools like Wireshark to decrypt captured TLS traffic after the fact. files︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
CAPEv2 now parses TLS session keys via parse_tlsdump_line in utils/tls.py, supporting NSSNSSNetwork Security Services, a set of cryptographic libraries maintained by Mozilla that implement TLS, PKI, and related standards; cyber tools use it to get a battle-tested, FIPS-validated crypto layer without writing their own. keylog labels (CLIENT_RANDOM, CLIENT_EARLY_TRAFFIC_SECRET, CLIENT_HANDSHAKE_TRAFFIC_SECRET, SERVER_HANDSHAKE_TRAFFIC_SECRET, CLIENT_TRAFFIC_SECRET_0, SERVER_TRAFFIC_SECRET_0, EARLY_EXPORTER_SECRET, EXPORTER_SECRET) alongside the existing Cape-native tlsdump.log format, and filters out all-zero secrets automatically.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
CAPEv2 adds tshark integration into modules/processing/network.py for network traffic processing, pulling in subprocess-based dissection and deduplication helpers deduplicate_file_to_storage and get_files_storage_path.
Finite State Platform is a software supply chain security tool that analyzes firmware and embedded software for vulnerabilities, risks, and compliance issues.
Finite State Platform's latest release adds KEVKEVCISA's Known Exploited Vulnerabilities catalog, a list of vulnerabilities with evidence of exploitation in real attacks. It helps teams prioritize fixes for problems attackers already use. membership flags on CVE update feeds, severity-ranked sorting for findings across five endpoints, match evidence for dependency-sourced BOMBOMA Bill of Materials listing every component, library, and dependency in a software project; cyber tools use it to track what is in a codebase and flag known-vulnerable components quickly. components, and a breaking change to how CWECWECommon Weakness Enumeration, a catalog of recurring software and hardware weakness types. It describes classes of mistakes, such as injection, rather than individual vulnerability instances. identifiers are formatted in CVE metadata responses.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
└──▷ WHAT SHIPPED ·
4 FEATURES0 OF 4 FEATURESmost completely described firstwhat's the number?
Each feature carries 0–100 for how completely the vendor documented
it — not how big or important the work is. A major capability
described in eight words scores low, and that is the finding.
depth0–40what it does and how it works or what changed
specificity0–30names real surfaces — APIs, flags, formats, limits, numbers
actionability0–30enough to go use it — a named endpoint, flag, or config key tops this; a UI path is a starting point
Code and config rank first by construction: a documented endpoint,
flag or config key scores at the top of specificity and actionability,
so it sorts above a feature described only in prose. Under
60 the notes go thin — everything below the “thinner
coverage” line is thinner documentation, not smaller work.
Hover any meter for that feature's three sub-scores.
0100
KEV membership flags on CVE updatesnew90how completely this was documenteddepth30/40specificity30/30actionability30/3090 / 100
New optional kev and vcKev boolean fields appear on oldValue and newValue in items returned by GET /v0/cves/updates, letting callers filter for CVE updates where the CVE was newly added to a Known Exploited Vulnerability catalog; such updates are served under updateType=exploit_update.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
Poll for CVE updates and check whether a change also reflects a new Known Exploited Vulnerability catalog addition, without a separate lookup.
Severity-ranked sorting for findings endpointsimproved90how completely this was documenteddepth30/40specificity30/30actionability30/3090 / 100
Version-scoped findings can now be sorted by severity rank (Critical, High, Medium, Low, Info, None, Unknown) and by component group and name via GET /v0/findings, GET /v0/projects/{id}/findings (up to 10 versions), GET /v0/scans/{id}/findings, GET /v0/components/{id}/findings, and GET /v0/versions/{id}/findings.
Retrieve version-scoped findings already sorted by severity rank to feed a triage dashboard without post-processing.
CWE identifiers returned as formatted stringsbreaking90how completely this was documenteddepth35/40specificity30/30actionability25/3090 / 100
GET /api/public/v0/cves/{cveId}/metadata and GET /api/public/v0/findings/{projectVersionId}/{findingId}/cves now return every weaknesses[].description[].value as a CWE-<number> string instead of a bare number (e.g. 399 becomes 'CWE-399'); placeholder values CWE-0 and CWE--1 are omitted, so a CVE whose only weakness was a placeholder now returns an empty weaknesses list.
Confirm a CVE weakness is now returned as a CWE-<number> string so downstream parsers that expected a bare integer can be updated.
Match evidence for dependency-sourced BOM componentsnew30how completely this was documenteddepth15/40specificity10/30actionability5/3030 / 100
Finite State Platform now shows match evidence, guidance, and dependencies for a component that comes from a project dependency in a product's Bill of MaterialsBill of MaterialsA structured list of every component, library, and dependency that makes up a piece of software or hardware, used by cyber tools to track what is inside a product and surface known vulnerabilities in its parts..
`GET /api/public/v0/cves/{cveId}/metadata` and `GET /api/public/v0/findings/{projectVersionId}/{findingId}/cves` now return every `weaknesses[].description[].value` as a `CWE-<number>` string instead of a bare number (e.g. `399` becomes `'CWE-399'`); the placeholders `CWE-0` and `CWE--1` are omitted, and a CVE whose only weakness was a placeholder now returns an empty `weaknesses` list.Our best guess
Fickling is a Python library that manipulates, analyzes, and deserializes pickle files, helping detect and create malicious pickle payloads for security research.
CVE-2026-22612: Detection bypass due to builtins blindness (CVSSCVSSCommon Vulnerability Scoring System, a method for describing a vulnerability's characteristics and severity. Its score is not the probability that attackers will exploit it. 7.8, affects <= v0.1.6)︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
3 more releases in this issue
· 2025-12-15 → 2026-01-09
CVE-2026-22606: Bypass via runpy.run_path() and runpy.run_module() (CVSSCVSSCommon Vulnerability Scoring System, a method for describing a vulnerability's characteristics and severity. Its score is not the probability that attackers will exploit it. 7.8, affects <= v0.1.6)︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
eCapture v2.11.0 adds richer process context to captured traffic and switches TLS modules to pcapngpcapngA file format for storing captured network packets, succeeding the older libpcap (.pcap) format; it supports multiple interfaces, metadata annotations, and timestamps with nanosecond precision in a single capture file. by default for out-of-the-box HTTP reconstruction.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
eCapture is an eBPF-based tool that captures plaintext network traffic, SSL/TLS data, and bash commands without requiring CA certificates or code modifications.
eCapture v2.11.0 adds richer process context to captured traffic and switches TLS modules to pcapngpcapngA file format for storing captured network packets, succeeding the older libpcap (.pcap) format; it supports multiple interfaces, metadata annotations, and timestamps with nanosecond precision in a single capture file. by default for out-of-the-box HTTP reconstruction.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
└──▷ WHAT SHIPPED ·
2 FEATURES0 OF 2 FEATURESmost completely described firstwhat's the number?
Each feature carries 0–100 for how completely the vendor documented
it — not how big or important the work is. A major capability
described in eight words scores low, and that is the finding.
depth0–40what it does and how it works or what changed
specificity0–30names real surfaces — APIs, flags, formats, limits, numbers
actionability0–30enough to go use it — a named endpoint, flag, or config key tops this; a UI path is a starting point
Code and config rank first by construction: a documented endpoint,
flag or config key scores at the top of specificity and actionability,
so it sorts above a feature described only in prose. Under
60 the notes go thin — everything below the “thinner
coverage” line is thinner documentation, not smaller work.
Hover any meter for that feature's three sub-scores.
0100
Pcapng default for TLS text-mode outputbreaking70how completely this was documenteddepth30/40specificity25/30actionability15/3070 / 100
The default text-mode output source for the tls, gotls, and gnutls modules is now pcapng instead of the previous default, so reconstructed HTTP messages are the out-of-the-box experience; existing scripts or workflows that relied on the old default text source may behave differently.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
— Clear before/after and named modules, but no migration command givenv2.11.0
0200
Process identity captured and exposed in pcapng/ecaptureqnew65how completely this was documenteddepth25/40specificity25/30actionability15/3065 / 100
eCapture now captures up to 256 bytes of process arguments per packet in pcapng output, giving analysts fuller context on which process generated each record. It also preserves the PID and Comm (process name) fields through the pcapng decryption pipeline so decrypted traffic retains its originating process identity, and the ecaptureq query interface now exposes these process arguments alongside TLS handshake metadata.
— Names fields and interface but no exact commands or query syntaxv2.11.0
└──▷ BREAKING ON UPGRADE
!
The default text-mode source for the `tls`, `gotls`, and `gnutls` modules is now pcapng instead of the previous default; existing scripts or workflows that relied on the old default text source may behave differently.Our best guess
Ettercap is an open-source network security tool for man-in-the-middle attacks, enabling live packet sniffing, filtering, and protocol analysis on local networks.
CVE-2026-125406: Stack buffer underflow in the telnet dissector's backspace handling (CVSSCVSSCommon Vulnerability Scoring System, a method for describing a vulnerability's characteristics and severity. Its score is not the probability that attackers will exploit it. 8.8, affects <= 0.8.4.1)︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
CVE-2026-125405: Heap out-of-bounds read in the HTTP dissector's NTLMNTLMA Windows authentication protocol that proves knowledge of a password through a challenge and response. Captured authentication exchanges can be abused in relay attacks when protections are absent. Type 3 parser (CVSSCVSSCommon Vulnerability Scoring System, a method for describing a vulnerability's characteristics and severity. Its score is not the probability that attackers will exploit it. 8.1, affects <= 0.8.4.1)︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
CVE-2026-125404: size_t underflow in ccmp_decrypt leads to unbounded stack write (CVSSCVSSCommon Vulnerability Scoring System, a method for describing a vulnerability's characteristics and severity. Its score is not the probability that attackers will exploit it. 7.5, affects <= 0.8.4.1)︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
CVE-2026-125402: Heap out-of-bounds read parsing the TLS SNISNIA TLS extension that lets a client declare which hostname it is connecting to during the handshake, allowing one server to host multiple certificates on a single IP address. extension during SSLSSLA cryptographic protocol for encrypting network connections, now deprecated in favor of TLS but whose name persists colloquially. Cyber tools reference it when handling HTTPS or other encrypted-channel inspection and termination. MITM (CVSSCVSSCommon Vulnerability Scoring System, a method for describing a vulnerability's characteristics and severity. Its score is not the probability that attackers will exploit it. 8.1, affects <= 0.8.4.1)︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
CVE-2026-125403: Out-of-bounds read and unchecked memmem result in sslw_remove_sts (CVSSCVSSCommon Vulnerability Scoring System, a method for describing a vulnerability's characteristics and severity. Its score is not the probability that attackers will exploit it. 8.1, affects <= 0.8.4.1)︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
CVE-2026-125401: Heap buffer overflow in the Gadu-Gadu dissector (gg_get_version) (CVSSCVSSCommon Vulnerability Scoring System, a method for describing a vulnerability's characteristics and severity. Its score is not the probability that attackers will exploit it. 8.1, affects <= 0.8.4.1)︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
CVE-2026-125400: Use-after-free in PCRE2 filter substitution (func_pcre) (CVSSCVSSCommon Vulnerability Scoring System, a method for describing a vulnerability's characteristics and severity. Its score is not the probability that attackers will exploit it. 8.8, affects <= 0.8.4.1)︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
Rspamd 4.3.0 ships new detection rules for DKIMDKIMAn email authentication standard that lets a sending mail server cryptographically sign outgoing messages so receiving servers can verify the signature was not forged or tampered with in transit.-signer abuse, tracking pixels, and far-future dates, adds SMTPUTF8SMTPUTF8An SMTP extension (RFC 6531) that allows UTF-8 encoded characters in email addresses and headers, enabling internationalised domain and mailbox names that ASCII-only SMTP cannot handle. address syntax checks, and expands archive and truncated-content inspection — alongside a long list of breaking changes to SPFSPFAn email authentication standard that lets a domain owner publish, via DNS, which mail servers are authorized to send on its behalf, so receiving servers can reject spoofed senders. evaluation order, DKIM validation, symbol option encoding, and several Lua API functions.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
Rspamd is an open-source spam filtering system that analyzes emails using rules, statistical methods, and machine learning to assign spam scores.
Rspamd 4.3.0 ships new detection rules for DKIMDKIMAn email authentication standard that lets a sending mail server cryptographically sign outgoing messages so receiving servers can verify the signature was not forged or tampered with in transit.-signer abuse, tracking pixels, and far-future dates, adds SMTPUTF8SMTPUTF8An SMTP extension (RFC 6531) that allows UTF-8 encoded characters in email addresses and headers, enabling internationalised domain and mailbox names that ASCII-only SMTP cannot handle. address syntax checks, and expands archive and truncated-content inspection — alongside a long list of breaking changes to SPFSPFAn email authentication standard that lets a domain owner publish, via DNS, which mail servers are authorized to send on its behalf, so receiving servers can reject spoofed senders. evaluation order, DKIM validation, symbol option encoding, and several Lua API functions.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
└──▷ WHAT SHIPPED ·
13 FEATURES3 OF 13 FEATURESmost completely described firstwhat's the number?
Each feature carries 0–100 for how completely the vendor documented
it — not how big or important the work is. A major capability
described in eight words scores low, and that is the finding.
depth0–40what it does and how it works or what changed
specificity0–30names real surfaces — APIs, flags, formats, limits, numbers
actionability0–30enough to go use it — a named endpoint, flag, or config key tops this; a UI path is a starting point
Code and config rank first by construction: a documented endpoint,
flag or config key scores at the top of specificity and actionability,
so it sorts above a feature described only in prose. Under
60 the notes go thin — everything below the “thinner
coverage” line is thinner documentation, not smaller work.
Hover any meter for that feature's three sub-scores.
0100
Archive inspection improvementsnew81how completely this was documenteddepth25/40specificity28/30actionability28/3081 / 100
archive.unpack gains new ignore_case and filters = false options to control unpacking behavior.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄ EXE_IN_ARCHIVE now also checks executable names inside archives named as documents, inside OOXMLOOXMLAn ISO-standardized file format family for Office documents (Word, Excel, PowerPoint), maintained by Microsoft, that stores content as XML inside a ZIP archive, giving cyber tools a parseable, scriptable structure for document analysis./ODF packages, and across the first 1000 archive entries.
Unpack an archive while ignoring case in filenames and disabling built-in libarchive filters, for controlled inspection of attachment contents.
— Runnable Lua call with exact option names given4.3.0
0201
SMTPUTF8 address syntax checksbreaking78how completely this was documenteddepth30/40specificity28/30actionability20/3078 / 100
Rspamd now validates SMTPUTF8 addresses per draft-ietf-mailmaint-smtputf8-syntax, scoring ASCII-syntax violations as INVALID_FROM_ASCII_ADDR and INVALID_RCPT_ASCII_ADDR (weight 1.0). The existing INVALID_FROM_8BIT and INVALID_RCPT_8BIT symbols now fire only for genuinely 8-bit addresses, so existing rules or score overrides targeting the old symbols need updating.
— Names exact symbols and spec but no config example4.3.0
0300
Truncated and incomplete content detectionnew75how completely this was documenteddepth30/40specificity30/30actionability15/3075 / 100
New symbols MIME_UNREADABLE_ARCHIVE, MIME_TRUNCATED_ARCHIVE, TEXT_WORDS_TRUNCATED, OOXML_INCOMPLETE, SVG_INCOMPLETE, PDF_NO_OBJECTS, PDF_DECODE_LIMIT, PDF_UNDECODED_STREAMS, PDF_PARSE_ERRORS, and PDF_TEXT_TRUNCATED report content that was cut short, incomplete, or unreadable across MIME, OOXML, SVG, and PDF documents. Text part get_stats also gains words_sampled and words_truncated fields to back this detection.
— All ten symbols and two API fields named verbatim4.3.0
0400
Symbol option encoding changebreaking75how completely this was documenteddepth30/40specificity30/30actionability15/3075 / 100
Symbol options are now stored verbatim; the characters []{};, and % are percent-encoded only when options are joined, affecting X-Spamd-Result, X-Virus, metadata_exporter, ClickHouse output, and the symbols log line. This changes the encoding of previously written option values.
— All affected surfaces named, no migration tool given4.3.0
0500
Lua API behavior changesbreaking73how completely this was documenteddepth30/40specificity28/30actionability15/3073 / 100
lua_util inflate now returns errors as a second value instead of a single return. lua_url get_redirected is no longer an alias of get_phished, so code relying on it will silently change behavior. The image flag on URLs is now set only when a URL is seen solely as an image source, changing what get_urls returns when iterating. A get_date result of 0 now means no date in the time('message') selector and the spamassassin plugin, so DATE_IN_PAST rules no longer match mail without a Date header.
— Four named Lua functions and exact behavior changes4.3.0
0602
SPF term evaluation order changebreaking70how completely this was documenteddepth30/40specificity25/30actionability15/3070 / 100
SPF records are now evaluated in the order of their terms, and the qualifier of an include or redirect is applied after the mechanisms. Records like -ip4:1.2.3.4 +ip4:1.2.3.0/24 -all that previously allowed an address to pass will now reject it.
DKIM signature validation hardeningbreaking65how completely this was documenteddepth30/40specificity25/30actionability10/3065 / 100
DKIM signatures with duplicated tags or l=0 over a body are now rejected, as are RSARSAThe oldest widely used public-key algorithm, for both encryption and signatures, with security resting on factoring large numbers. A large quantum computer would break it. keys under 1024 bits and keys whose type does not match the k= tag. The d= domain is now lowercased consistently in results, symbol options, and glob maps.
— Clear before/after but no migration command4.3.0
0800
BAD_DKIM_SIGNER rulenew60how completely this was documenteddepth20/40specificity25/30actionability15/3060 / 100
New multimap rule BAD_DKIM_SIGNER matches every checked DKIM d= signer domain against the bad_dkim_signers glob map.
Archive output and file-counting limitsbreaking60how completely this was documenteddepth25/40specificity25/30actionability10/3060 / 100
get_files(n) and get_files_full(n) now count files rather than names, with other names of an entry flagged as aliases. archive.unpack with no explicit limit is now capped at 1 GiB of output and one million headers.
— Named functions and limits but no migration snippet4.3.0
1000
Statistics dump filtering by usernew60how completely this was documenteddepth15/40specificity20/30actionability25/3060 / 100
The statistics_dump command gains a new dump --user option to filter statistics dumps by user.
HTML tracking pixel detectionnew50how completely this was documenteddepth20/40specificity20/30actionability10/3050 / 100
New rule HTML_TINY_PIXEL_TRACKER and composite rule HTML_TINY_PIXEL_PROBE detect contentless messages carrying tiny external tracking pixels.
— Names rules, no scoring or map detail given4.3.0
1200
Phishing URL mismatch rewrite trust mapsbreaking45how completely this was documenteddepth20/40specificity15/30actionability10/3045 / 100
Phishing detection adds explicit rewrite trust maps for URL mismatch scoring. URL mismatches are now scored separately from phishing, with rewrite permissions sourced from dynamic maps, so existing phishing score tuning may need adjustment.
URL mismatches are now scored separately from phishing; rewrite permissions come from dynamic maps, so existing phishing score tuning may need adjustment.Our best guess
Symbol options are now stored verbatim; `[]{};,` and `%` are percent-encoded only when options are joined (X-Spamd-Result, X-Virus, metadata_exporter, ClickHouse, the symbols log line), which changes the encoding of previously written option values.Our best guess
SPF records are now evaluated in the order of their terms, and the qualifier of an `include` or `redirect` is applied after the mechanisms — records like `-ip4:1.2.3.4 +ip4:1.2.3.0/24 -all` that previously allowed an address to pass will now reject it.Our best guess
DKIM signatures with duplicated tags or `l=0` over a body are now rejected, as are RSA keys under 1024 bits and keys whose type does not match the `k=` tag. The `d=` domain is lowercased in results, symbol options, and glob maps.Our best guess
SMTPUTF8 ASCII syntax violations now score as `INVALID_FROM_ASCII_ADDR` and `INVALID_RCPT_ASCII_ADDR` (1.0); `INVALID_FROM_8BIT` and `INVALID_RCPT_8BIT` fire only for 8-bit addresses — existing rules or score overrides targeting the old symbols will need updating.Our best guess
get_files(n) and get_files_full(n) now count files rather than names; other names of an entry are flagged as aliases. `archive.unpack` with no limit is now capped at 1 GiB of output and one million headers.Our best guess
The image flag on URLs is now set only when a URL is seen solely as an image source; `get_urls` now returns a link that is also an image source, changing iteration results.Our best guess
A `get_date` result of `0` now means 'no date' in the time('message') selector and the spamassassin plugin, so `DATE_IN_PAST` rules no longer match mail without a `Date` header.Our best guess
Maltrail is a malicious traffic detection system that uses blacklists and heuristics to identify suspicious network activity and potential threats.
Maltrail 3.6 adds ERSPANERSPANA Cisco-originated GRE-based tunneling protocol that encapsulates mirrored network traffic and ships it across Layer 3 to a remote capture point, letting cyber tools analyze traffic from distant segments without local tap hardware. and MPLSMPLSA packet-forwarding method that routes traffic using short numeric labels rather than full IP lookups, letting network operators engineer paths and stack tunnels with lower per-hop overhead. packet decapsulation, new DNS CNAMECNAMEA DNS record type that maps one hostname to another as a canonical alias, letting cyber tools track infrastructure by following chains of names to their true origin., ICMP-tunnelingICMP-tunnelingA covert channel that encodes arbitrary data inside ICMP echo request/reply packets, exploiting firewalls that permit ping traffic to smuggle command-and-control or exfiltrated data past network controls. and IDNIDNInternationalized Domain Names — domain names containing non-ASCII characters encoded via Punycode so DNS can handle them; cyber tools flag them because they enable homograph phishing attacks using visually similar Unicode characters.-homograph detection heuristics, a trail-store memory-map cache that cuts cold reloads from 0.78s to 0.08s, and a new sensor health and alert-persistence layer, alongside breaking config key renames in maltrail.conf.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
└──▷ WHAT SHIPPED ·
10 FEATURES1 OF 10 FEATURESmost completely described firstwhat's the number?
Each feature carries 0–100 for how completely the vendor documented
it — not how big or important the work is. A major capability
described in eight words scores low, and that is the finding.
depth0–40what it does and how it works or what changed
specificity0–30names real surfaces — APIs, flags, formats, limits, numbers
actionability0–30enough to go use it — a named endpoint, flag, or config key tops this; a UI path is a starting point
Code and config rank first by construction: a documented endpoint,
flag or config key scores at the top of specificity and actionability,
so it sorts above a feature described only in prose. Under
60 the notes go thin — everything below the “thinner
coverage” line is thinner documentation, not smaller work.
Hover any meter for that feature's three sub-scores.
0100
Trail-store memory-map cachenew95how completely this was documenteddepth35/40specificity30/30actionability30/3095 / 100
A new memory-map cache for the trail table can be enabled or disabled with TRAIL_STORE_CACHE in maltrail.conf, and its integrity can be verified with maltrail-sensor --verify-trail-store.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄ On 1.9M trails it reduces a cold -T reload from 0.78s to 0.08s and peak RSS from 145MB to 89MB.
Verify the trail-store cache is consistent with the current trails.csv before relying on the fast mapped-load path.
$ maltrail-sensor --verify-trail-store
— Named config key, CLI flag and measured before/after numbers3.6
0200
Sensor health monitoring and dashboard reportingnew90how completely this was documenteddepth35/40specificity30/30actionability25/3090 / 100
Adds sensor health and alert-persistence infrastructure: a configurable stale threshold via SENSOR_STALE_AFTER in maltrail.conf (default 15 minutes) with stale/recovery alerts logged to error.log; a configurable remote clock-skew tolerance via LOG_SERVER_SKEW in maltrail.conf that counts and shows out-of-window senders instead of silently misfiling them; a heartbeat file at LOG_DIR/<sensor>.heartbeat for single-host installs without LOG_SERVER; persistent sensor state in LOG_DIR/sensors.json so a server restart retains knowledge of which sensors should be reporting; a persistent alert queue in LOG_DIR/alerts.json preserving up to 1,000 pending alerts and the tail position across restarts; and the /sensors endpoint now also shows webhook delivery status and, sourced from packet counts carried in the heartbeat, dropped-packet share per sensor, with drop alerting tunable per sensor via SENSOR_DROP_ALERT and SENSOR_DROP_WINDOW in maltrail.conf.
Tune drop alerting on a busy sensor so the server notifies once per hour when more than 1% of packets are dropped.
— Five named config keys and file paths, no single runnable command Names config keys with a concrete tuning example3.6
0300
Typed TLS fingerprint trails and data validation toolingnew85how completely this was documenteddepth30/40specificity30/30actionability25/3085 / 100
Static trail files now support typed TLS fingerprints per line (ja3:<md5>, ja4:<ja4>, cert:<sha1 or sha256>), with check_trails.py validating each type and certificate trails accepting SHA-256SHA-256A cryptographic hash function that turns data into a fixed 256-bit fingerprint. It is used to check integrity; hashing does not encrypt the original data. as well as SHA-1SHA-1A cryptographic hash function standardized by NIST that produces a 160-bit digest; tools use it for file integrity checks, though it is now considered cryptographically broken and unsuitable for collision-resistant applications.. A new format gate sensor/tools/check_data.py validates every data/*.txt file, catching loader-silent typos, unsupported inline comments, and IPv6IPv6The current version of the Internet Protocol, standardized by the IETF, using 128-bit addresses to replace the exhausted IPv4 space. Cyber tools target it because modern networks run dual-stack or IPv6-only. ranges, to be run on any data/ change.
— Names formats and both validator scripts to run3.6
0401
Config key renames and capture defaults in maltrail.confbreaking85how completely this was documenteddepth30/40specificity30/30actionability25/3085 / 100
CAPTURE_FANOUT is renamed to CAPTURE_WORKERS and CHECK_TLS_CERTIFICATES is renamed to CHECK_TLS_FINGERPRINTS in maltrail.conf; old names are still read but should be updated, and PROCESS_COUNT is no longer set. CAPTURE_BUFFER is no longer mandatory — the default 10% libpcap ring is used if unset, though existing explicit values are kept. The default CAPTURE_FILTER now restricts QUICQUICA UDP-based transport protocol standardized by the IETF that bakes in TLS 1.3, multiplexed streams, and connection migration, giving cyber tools low-latency encrypted transport without head-of-line blocking. to long-header (handshake) packets only, while existing custom filters are kept until manually updated.
— Names every renamed/changed config key with migration behaviour3.6
0500
Configurable alert throttleimproved73how completely this was documenteddepth25/40specificity28/30actionability20/3073 / 100
Alert throttle is now configurable via ALERT_THROTTLE in maltrail.conf (default 86400, one message per threat per day); the %(threat)s template variable now carries the dashboard threat ID.
— Named config key and default, no example command given3.6
0600
Configurable dashboard session expirationnew70how completely this was documenteddepth20/40specificity25/30actionability25/3070 / 100
A new SESSION_EXPIRATION_HOURS config key in maltrail.conf controls how long a dashboard session stays valid, defaulting to 24 hours.
Reduce dashboard session length to 8 hours for analyst accounts on shared workstations.
ini
SESSION_EXPIRATION_HOURS = 8
— Named config key with default and a runnable example3.6
0700
Dashboard day-loading via Web Workerimproved70how completely this was documenteddepth25/40specificity25/30actionability20/3070 / 100
Dashboard day-loading now runs in a Web Worker (js/worker.js) against the same engine as the main page (js/engine.js), keeping the table interactive during load; the in-thread path can be forced with localStorage.mt_worker = "0".
— Names files and the override flag to disable it3.6
0800
ERSPAN and MPLS packet decapsulationnew68how completely this was documenteddepth30/40specificity28/30actionability10/3068 / 100
The sensor now decapsulates ERSPAN type I and type III (including the optional platform sub-header), MPLS label stacks up to 8 deep, and Ethernet over MPLS with or without a control word.
— Mechanism detailed but no config surface to act on3.6
0900
New detection heuristicsnew68how completely this was documenteddepth30/40specificity28/30actionability10/3068 / 100
DNS responses are now checked against trail tables for listed CNAME targets, naming both the queried and aliased domain when a client resolves an innocent name aliased to a listed one; a new icmp_tunneling heuristic detects ICMP-based tunnels (icmpsh, ptunnel, exfiltration) by flagging more than 10 distinct echo payloads from one host to one destination that match no known ping-tool pattern; and a new idn_homograph heuristic detects IDN homograph attacks where an xn-- label with Cyrillic, Greek, or Armenian lookalike letters resolves to a whitelisted domain it is not.
— Clear behaviour described, no tunable surface named Names heuristic and its exact trigger threshold Names heuristic and the scripts it checks3.6
thinner coverage below
1000
Reverse DNS lookup action in dashboardnew45how completely this was documenteddepth15/40specificity15/30actionability15/3045 / 100
The dashboard adds a right-click Reverse DNS action on event addresses, resolved by the server on demand.
`CAPTURE_FANOUT` is renamed to `CAPTURE_WORKERS` and `CHECK_TLS_CERTIFICATES` is renamed to `CHECK_TLS_FINGERPRINTS` in `maltrail.conf`; old names are still read but should be updated. `PROCESS_COUNT` is no longer set.Our best guess
`CAPTURE_BUFFER` is no longer mandatory in `maltrail.conf`; the default 10% libpcap ring is used if unset, so existing configs that set it explicitly keep their value unchanged but new installs omit it.Our best guess
An existing `maltrail.conf` with a custom `CAPTURE_FILTER` keeps its current filter until manually updated; the new default filter restricts QUIC to long-header (handshake) packets only.Our best guess
Nudge Security is a SaaS security platform that discovers shadow IT assets, manages identity governance, and hardens security posture.
Nudge Security adds breached password detection with in-browser remediation nudges and AI agent verdict filtering for browser extensions.
New detectionConfigure breached password detection via browser extension settings to detect when employees log in with passwords exposed in known data breaches, trigger in-browser nudges to prompt password changes, and surface affected apps and accounts on a new password risk findings card in the Identity & location dashboard.
Filter the browser extension table by AI agent verdict — malicious, compromised, risky, or safe — to compare verdicts against each extension's approval status, such as finding malicious extensions that are still approved.
Sysdig Secure's Registry Scanner (v0.14.0) adds credential-free authentication to Azure Container Registry via managed or workload identityworkload identityAn identity assigned to software, such as a service or container, rather than a person. It lets the workload authenticate and receive permissions without sharing a user's credentials. and extends registry scanning to Red Hat OpenShift Service on AWS (ROSA) with hosted control planes reading Amazon ECRECRAmazon Elastic Container Registry, AWS's managed service for storing and distributing Docker and OCI container images, giving cyber tools a private, IAM-gated registry without self-hosting infrastructure..︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
Sysdig Secure is a cloud-native security platform that detects threats, manages vulnerabilities, and ensures compliance across containers, Kubernetes, and cloud environments.
Sysdig Secure's Registry Scanner (v0.14.0) adds credential-free authentication to Azure Container Registry via managed or workload identityworkload identityAn identity assigned to software, such as a service or container, rather than a person. It lets the workload authenticate and receive permissions without sharing a user's credentials. and extends registry scanning to Red Hat OpenShift Service on AWS (ROSA) with hosted control planes reading Amazon ECRECRAmazon Elastic Container Registry, AWS's managed service for storing and distributing Docker and OCI container images, giving cyber tools a private, IAM-gated registry without self-hosting infrastructure..︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
└──▷ WHAT SHIPPED ·
2 FEATURES2 OF 2 FEATURESmost completely described firstwhat's the number?
Each feature carries 0–100 for how completely the vendor documented
it — not how big or important the work is. A major capability
described in eight words scores low, and that is the finding.
depth0–40what it does and how it works or what changed
specificity0–30names real surfaces — APIs, flags, formats, limits, numbers
actionability0–30enough to go use it — a named endpoint, flag, or config key tops this; a UI path is a starting point
Code and config rank first by construction: a documented endpoint,
flag or config key scores at the top of specificity and actionability,
so it sorts above a feature described only in prose. Under
60 the notes go thin — everything below the “thinner
coverage” line is thinner documentation, not smaller work.
Hover any meter for that feature's three sub-scores.
0101
ACR managed/workload identity authentication in Registry Scannernew77how completely this was documenteddepth30/40specificity27/30actionability20/3077 / 100
Registry Scanner v0.14.0 authenticates with Azure Container Registry (ACR) using managed identity or workload identity when running on AKSAKSAzure Kubernetes Service, Microsoft's managed Kubernetes offering, which lets cyber tools offload cluster provisioning and scaling to Azure so engineers focus on workload logic instead of control-plane operations. — via the cluster's kubelet identity or a user-assigned identity federated to its service account — removing the need to store registry credentials.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄ Both RBACRBACRole-based access control, where permissions are assigned to roles such as reader or administrator and users receive those roles. It makes access easier to manage than individual permission lists. and RBAC + ABACABACAttribute-based access control, where rules evaluate facts about the user, resource and request. Access can depend on conditions such as department, data sensitivity or device status. repository permission modes are supported, and new log messages distinguish ACR identity authentication failures between an identity that cannot get a token, one with no role on the registry, and a request blocked by registry network rules.
— Names mechanism, permission modes and failure cases but no config keys or flags.changelog-20261114-348548da
0202
ROSA hosted control plane support in Registry Scannernew62how completely this was documenteddepth25/40specificity22/30actionability15/3062 / 100
Registry Scanner v0.14.0 now supports Red Hat OpenShift Service on AWS (ROSA) with hosted control planes, reading Amazon Elastic Container Registry (ECR) via an IAMIAMIdentity and Access Management — a framework of policies and services for controlling who can authenticate to systems and what resources they are authorized to use, letting cyber tools enforce least-privilege and audit access. role bound to its service account, so AWS access keys no longer need to be stored in the cluster.
Steampipe v2.5.0 upgrades the embedded PostgreSQL to 14.24.0 and tightens its authentication model, relocating the unix socket and locking down root access and pg_hba.conf handling.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
Steampipe is an open-source tool that lets you query cloud infrastructure, APIs, and security data using SQL.
Steampipe v2.5.0 upgrades the embedded PostgreSQL to 14.24.0 and tightens its authentication model, relocating the unix socket and locking down root access and pg_hba.conf handling.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
└──▷ WHAT SHIPPED ·
5 FEATURES0 OF 5 FEATURESmost completely described firstwhat's the number?
Each feature carries 0–100 for how completely the vendor documented
it — not how big or important the work is. A major capability
described in eight words scores low, and that is the finding.
depth0–40what it does and how it works or what changed
specificity0–30names real surfaces — APIs, flags, formats, limits, numbers
actionability0–30enough to go use it — a named endpoint, flag, or config key tops this; a UI path is a starting point
Code and config rank first by construction: a documented endpoint,
flag or config key scores at the top of specificity and actionability,
so it sorts above a feature described only in prose. Under
60 the notes go thin — everything below the “thinner
coverage” line is thinner documentation, not smaller work.
Hover any meter for that feature's three sub-scores.
0100
pg_hba.conf management hardenedbreaking75how completely this was documenteddepth30/40specificity20/30actionability25/3075 / 100
Custom host-based authentication rules can be placed in pg_hba.conf.d to be applied first on every start.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄ Steampipe now writes pg_hba.conf before every startup and never overwrites a file it did not create, stopping with an error if it encounters a pg_hba.conf it did not write, requiring manual resolution before startup proceeds.
— Names exact files and gives a concrete fallback path for custom rulesv2.5.0
0200
Root database access locked to unix socketbreaking65how completely this was documenteddepth25/40specificity20/30actionability20/3065 / 100
Steampipe now restricts root database connections to the OS user who launched it, via the unix socket only; any TCPTCPA network protocol that delivers a reliable, ordered stream of bytes between two endpoints. It detects lost data and retransmits it. or other-user root connections that previously worked will be rejected.
— Clear mechanism and scope, no example or migration command givenv2.5.0
0300
Unix socket relocated to data directorybreaking65how completely this was documenteddepth25/40specificity20/30actionability20/3065 / 100
The PostgreSQL unix socket now lives in the data directory instead of /tmp, with a clear error and guidance if the resulting path exceeds the OS limit; external clients or scripts that hardcode the /tmp socket path will fail to connect.
— Names old and new socket locations and the failure conditionv2.5.0
thinner coverage below
0400
PostgreSQL upgraded to 14.24.0 with data preservationimproved55how completely this was documenteddepth25/40specificity20/30actionability10/3055 / 100
Steampipe now upgrades the embedded PostgreSQL to 14.24.0, preserving the existing data folder on same-major upgrades and carrying over the newest older database.
— Describes upgrade mechanism but no user action requiredv2.5.0
0500
Plugin SDK and FDW dependency updatesimproved30how completely this was documenteddepth10/40specificity15/30actionability5/3030 / 100
Steampipe now updates steampipe-plugin-sdk to v6 and the embedded steampipe-postgres-fdw to v2.3.0.
— Bare version bump with no explanation of impactv2.5.0
└──▷ BREAKING ON UPGRADE
!
The PostgreSQL unix socket path has moved from `/tmp` to the data directory; external clients or scripts that hardcode the `/tmp` socket path will fail to connect.Our best guess
Only the OS user who runs Steampipe can now connect as `root`, and only through the unix socket; any TCP or other-user `root` connections that previously worked will be rejected.Our best guess
Steampipe now writes `pg_hba.conf` before every start and stops with an error if it encounters a `pg_hba.conf` it did not write; a manually managed `pg_hba.conf` will block startup until resolved.Our best guess
Plerion consolidates identity-related capabilities into a new Identity security section, adding a cross-cloud Overview and reorganizing access grant review into a single Access grants tab that replaces the old per-cloud overview tabs.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
Plerion is a cloud security platform that provides comprehensive risk management, vulnerability detection, and threat prioritization across cloud environments and workloads.
Plerion consolidates identity-related capabilities into a new Identity security section, adding a cross-cloud Overview and reorganizing access grant review into a single Access grants tab that replaces the old per-cloud overview tabs.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
└──▷ WHAT SHIPPED ·
2 FEATURES0 OF 2 FEATURESmost completely described firstwhat's the number?
Each feature carries 0–100 for how completely the vendor documented
it — not how big or important the work is. A major capability
described in eight words scores low, and that is the finding.
depth0–40what it does and how it works or what changed
specificity0–30names real surfaces — APIs, flags, formats, limits, numbers
actionability0–30enough to go use it — a named endpoint, flag, or config key tops this; a UI path is a starting point
Code and config rank first by construction: a documented endpoint,
flag or config key scores at the top of specificity and actionability,
so it sorts above a feature described only in prose. Under
60 the notes go thin — everything below the “thinner
coverage” line is thinner documentation, not smaller work.
Hover any meter for that feature's three sub-scores.
0100
Identity security Overview sectionnew75how completely this was documenteddepth30/40specificity25/30actionability20/3075 / 100
Plerion adds a new Identity security section at /identity-security that surfaces cross-cloud identity posture across AWS, Azure, and GCP.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄ The Overview tab shows account and identity counts, open findings with 30-day trends, a 7-day change summary, risk by account, and a prioritized findings list, with a cloud selector to filter by provider.
Review your cross-cloud identity posture and prioritized open findings in one place.
📍In the side navigation, go to Identity security › Overview. Use the cloud selector to filter to AWS, Azure, or GCP individually, then check Identity risk by account and the prioritized findings list.
Access grants tab with external exposure viewsbreaking75how completely this was documenteddepth30/40specificity25/30actionability20/3075 / 100
The Access grants tab, now at /identity-security/access-grants, lets users review resource access grants, with new External and Untrusted external views for scoping external access exposure. This retires the separate AWS, Azure, and GCP overview tabs, including the identity risk trend, identity top risks, and the AWS External access card, consolidating external access review into this single tab.
Investigate external access exposure across your cloud environments after the AWS External access card retirement.
📍In the side navigation, go to Identity security › Access grants, then select the External or Untrusted external view to scope and review externally accessible resource grants.
The separate AWS, Azure, and GCP overview tabs (with identity risk trend, identity top risks, and the AWS External access card) have been retired; external access is now reviewed via the External and Untrusted external views on the Access grants tab at `/identity-security/access-grants`.Our best guess
Rapid7 InsightCloudSec is a cloud-native security platform that provides continuous compliance, threat detection, and automated remediation across multi-cloud environments.
Rapid7 InsightCloudSec's November release expands cloud inventory coverage and adds new posture insights and search filters.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
└──▷ WHAT SHIPPED ·
1 FEATURE0 OF 1 FEATUREmost completely described firstwhat's the number?
Each feature carries 0–100 for how completely the vendor documented
it — not how big or important the work is. A major capability
described in eight words scores low, and that is the finding.
depth0–40what it does and how it works or what changed
specificity0–30names real surfaces — APIs, flags, formats, limits, numbers
actionability0–30enough to go use it — a named endpoint, flag, or config key tops this; a UI path is a starting point
Code and config rank first by construction: a documented endpoint,
flag or config key scores at the top of specificity and actionability,
so it sorts above a feature described only in prose. Under
60 the notes go thin — everything below the “thinner
coverage” line is thinner documentation, not smaller work.
Hover any meter for that feature's three sub-scores.
0100
New cloud resource coverage, Insights, and Query Filtersnew30how completely this was documenteddepth12/40specificity12/30actionability6/3030 / 100
InsightCloudSec now supports Amazon SageMaker Ground Truth resources for cloud inventory and analysis, and the release adds new Insights for cloud security posture checks along with new Query Filters for narrowing cloud resource searches.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
— Names one specific resource type but no concrete Insight or filter names given.2026-11-10-5f389534
Kubescape is an open-source Kubernetes security platform that scans clusters, YAML files, and Helm charts for misconfigurations and vulnerabilities against security frameworks.
Kubescape's v4.0.21 release focuses on scan customization and observability, adding ValidatingAdmissionPolicyValidatingAdmissionPolicyA Kubernetes mechanism that lets cluster operators enforce custom rules on API objects using CEL expressions, rejecting non-compliant requests without deploying a separate webhook server.-based custom rules, a file watcher for continuous scanning, and refined evidence output.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
└──▷ WHAT SHIPPED ·
7 FEATURES1 OF 7 FEATURESmost completely described firstwhat's the number?
Each feature carries 0–100 for how completely the vendor documented
it — not how big or important the work is. A major capability
described in eight words scores low, and that is the finding.
depth0–40what it does and how it works or what changed
specificity0–30names real surfaces — APIs, flags, formats, limits, numbers
actionability0–30enough to go use it — a named endpoint, flag, or config key tops this; a UI path is a starting point
Code and config rank first by construction: a documented endpoint,
flag or config key scores at the top of specificity and actionability,
so it sorts above a feature described only in prose. Under
60 the notes go thin — everything below the “thinner
coverage” line is thinner documentation, not smaller work.
Hover any meter for that feature's three sub-scores.
0101
ValidatingAdmissionPolicy support in custom rulesnew60how completely this was documenteddepth20/40specificity20/30actionability20/3060 / 100
kubescape scan now accepts user-authored ValidatingAdmissionPolicy YAML via the --custom-rules flag.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
— Names flag and command, clear usage path.v4.0.21
thinner coverage below
0200
Test case filtering and JUnit XML reportsnew55how completely this was documenteddepth20/40specificity15/30actionability20/3055 / 100
Adds --run to filter which policy test cases run, plus support for writing policy test results as a JUnit XMLJUnit XMLA widely adopted test-report file format, originally from the Java JUnit framework, that records test suite results as XML; cyber tools emit it so their output plugs into CI systems like Jenkins or GitLab without custom parsers. report.
— Names the flag and output format but little mechanism detail.v4.0.21
0300
Evidence output improvements for cluster scansimproved55how completely this was documenteddepth20/40specificity20/30actionability15/3055 / 100
Prints a kubectl command for each evidence path on cluster scans, and --show-evidence now shows current values for delete paths.
— Names the flag and output change but limited mechanism.v4.0.21
0400
File watcher for kubescape watchnew50how completely this was documenteddepth20/40specificity15/30actionability15/3050 / 100
Adds a file watcher with a debounced, sequential scan scheduler for kubescape watch.
— Describes scheduling mechanism and names the command.v4.0.21
0500
verify_remediation tool in MCP servernew40how completely this was documenteddepth15/40specificity15/30actionability10/3040 / 100
Adds a verify_remediation tool to the MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface. server for local IaCIaCInfrastructure as Code: the practice of defining and provisioning compute infrastructure through machine-readable config files rather than manual processes, giving cyber tools repeatable, version-controlled environment setup. remediation verification.
— Names the tool but no mechanism or usage detail.v4.0.21
0600
Kustomize resource attribution to source fileimproved35how completely this was documenteddepth15/40specificity10/30actionability10/3035 / 100
Attributes each Kustomize-rendered resource to the source file Kustomize read it from.
— States the behavior change but no further specifics.v4.0.21
0700
JSON schema for fleet reportnew25how completely this was documenteddepth10/40specificity10/30actionability5/3025 / 100
Trend Vision One is a unified cybersecurity platform that integrates threat detection, response, and risk management across endpoints, networks, and cloud environments.
Exaforce runs AI Exabots that detect, triage, investigate and respond to threats across cloud, SaaS, identity, endpoint and code on a real-time knowledge graph.
Exaforce launched Force ER, an on-demand emergency incident response service with a 15-minute responder connection and 24-hour environment onboarding, backed by automated log replay, a shared War Room, mobile incident controls, and an optional retainer tier.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
└──▷ WHAT SHIPPED ·
5 FEATURES4 OF 5 FEATURESmost completely described firstwhat's the number?
Each feature carries 0–100 for how completely the vendor documented
it — not how big or important the work is. A major capability
described in eight words scores low, and that is the finding.
depth0–40what it does and how it works or what changed
specificity0–30names real surfaces — APIs, flags, formats, limits, numbers
actionability0–30enough to go use it — a named endpoint, flag, or config key tops this; a UI path is a starting point
Code and config rank first by construction: a documented endpoint,
flag or config key scores at the top of specificity and actionability,
so it sorts above a feature described only in prose. Under
60 the notes go thin — everything below the “thinner
coverage” line is thinner documentation, not smaller work.
Hover any meter for that feature's three sub-scores.
0101
Force ER on-demand emergency incident responsenew80how completely this was documenteddepth25/40specificity25/30actionability30/3080 / 100
Exaforce adds Force ER, an emergency incident response service available to any organization under active attack with no existing customer relationship or retainer required.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄ It connects a live responder within 15 minutes and goes live in the customer's environment within 24 hours; the incident hotline is +1 (844) 660-1663 or exaforce.com/incident-response.
Automated 30-day log replay for incident timelinesnew55how completely this was documenteddepth25/40specificity20/30actionability10/3055 / 100
As soon as an organization connects to Force ER, Exaforce replays the last 30 days of identity, cloud, SaaSSaaSSoftware delivered over the internet and billed by subscription, hosted and operated by the vendor so the buyer skips installation and infrastructure management., code, and VPNVPNA virtual private network, an encrypted tunnel between devices or networks. It carries traffic across another network as though the endpoints had a private connection. logs to build a pre-built incident timeline without manual log correlation.
Mobile incident response via ExaGoimproved55how completely this was documenteddepth20/40specificity20/30actionability15/3055 / 100
Exaforce's ExaGo mobile interface now lets responders approve containment actions, ask incident-status questions, and join the War Room from a phone or Apple Watch during an active incident.
Shared War Room for incident collaborationnew45how completely this was documenteddepth20/40specificity15/30actionability10/3045 / 100
Force ER adds a shared War Room where the customer's team and the MDRMDRManaged detection and response, an outsourced service that monitors security signals, investigates threats and helps respond to incidents. It combines specialist analysts with detection tools. team work from the same live incident view, with communication synced to Slack and findings routed to named stakeholders.
Optional incident response retainer tiernew45how completely this was documenteddepth20/40specificity15/30actionability10/3045 / 100
Exaforce adds an optional incident response retainer tier that pre-connects data sources, establishes named contacts, and provides a contractual SLASLAA formal agreement between a service provider and customer that defines expected uptime, response times, and remedies for failures, giving cyber tools a contractual baseline for availability and support commitments. so investigation begins on the first call.
Tenzir's headline release this window is v7.0.0-rc.1, adding an MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface. server and built-in assistant for AI agents, versioned packages with live rollout, and breaking changes to several source operators, while companion tools pushed pushdown-based query acceleration for SQL/Splunk/SentinelOne sources, IcebergIcebergAn open table format for large analytic datasets, maintained by the Apache Software Foundation, that gives cyber tools reliable schema evolution, time-travel queries, and atomic commits on object storage. sink efficiency, Tenzir Test's move to Tenzir 7, and TQL editor tooling updates.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
Tenzir is a data pipeline engine for security teams that collects, transforms, enriches, and routes telemetry.
Tenzir's headline release this window is v7.0.0-rc.1, adding an MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface. server and built-in assistant for AI agents, versioned packages with live rollout, and breaking changes to several source operators, while companion tools pushed pushdown-based query acceleration for SQL/Splunk/SentinelOne sources, IcebergIcebergAn open table format for large analytic datasets, maintained by the Apache Software Foundation, that gives cyber tools reliable schema evolution, time-travel queries, and atomic commits on object storage. sink efficiency, Tenzir Test's move to Tenzir 7, and TQL editor tooling updates.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
└──▷ WHAT SHIPPED ·
7 FEATURES11 FEATURES5 OF 11 FEATURESmost completely described firstwhat's the number?
Each feature carries 0–100 for how completely the vendor documented
it — not how big or important the work is. A major capability
described in eight words scores low, and that is the finding.
depth0–40what it does and how it works or what changed
specificity0–30names real surfaces — APIs, flags, formats, limits, numbers
actionability0–30enough to go use it — a named endpoint, flag, or config key tops this; a UI path is a starting point
Code and config rank first by construction: a documented endpoint,
flag or config key scores at the top of specificity and actionability,
so it sorts above a feature described only in prose. Under
60 the notes go thin — everything below the “thinner
coverage” line is thinner documentation, not smaller work.
Hover any meter for that feature's three sub-scores.
0101
Tenzir Test runs on Tenzir 7 via tenzir runnew80how completely this was documenteddepth30/40specificity25/30actionability25/3080 / 100
Tenzir Test v3 runs test pipelines using the tenzir run command from Tenzir 7, fetches a package's locked dependencies before running its tests, and automatically injects project fixtures into Python tests without requiring explicit imports.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
Query pushdown into DuckDB, ClickHouse, MySQL, MSSQL, and SplunkUpdated supported platformUpdated compliance frameworkNew integrationUpdated integrationNew compliance frameworkimproved80how completely this was documenteddepth35/40specificity30/30actionability15/3080 / 100
Tenzir now pushes filters, sorting, and limits after summaries down into SQL databases (DuckDB, ClickHouse, MySQL, Microsoft SQL Server) and Splunk, reducing the data returned over the wire. Splunk searches now return typed results rather than raw strings for faster downstream processing, and pushdown now reaches through more stateful operators, ocsf_cast, and field selection functions in pipelines.
Tenzir Test v3 breaking changes for Tenzir 6 projectsbreaking75how completely this was documenteddepth25/40specificity25/30actionability25/3075 / 100
The built-in node fixture is removed; projects still on Tenzir 6 should pin tenzir-test below version 3, since Tenzir Test v3 requires Tenzir 7 and will not work with Tenzir 6.
Native semantics for Hydrolix, Microsoft Graph, CloudWatch, Microsoft 365 Activity, and Velociraptor sourcesNew integrationbreaking75how completely this was documenteddepth30/40specificity30/30actionability15/3075 / 100
Operators placed after from_hydrolix, from_microsoft_graph, and from_amazon_cloudwatch now use native Hydrolix, Microsoft GraphMicrosoft GraphMicrosoft's unified REST API for accessing data and services across Microsoft 365, Azure AD, and related cloud products, giving cyber tools a single authenticated endpoint to query users, emails, files, and sign-in logs., and CloudWatch Logs semantics respectively, changing prior behavior. Operators after from_microsoft_365_activity now apply content-skipping filters and limits, and operators after from_velociraptor now stay in the pipeline by design rather than exiting early.
Automatic package.yaml version bump on releaseimproved50how completely this was documenteddepth20/40specificity20/30actionability10/3050 / 100
Tenzir Ship releases now automatically update the version field in a Tenzir package's package.yaml, removing the need for manual manifest edits after cutting a release.
Tree-sitter TQL grammar adds editor highlighting and structure queriesimproved50how completely this was documenteddepth20/40specificity20/30actionability10/3050 / 100
Tenzir's Tree-sitterTree-sitterA parser-generator library, maintained by the Tree-sitter project, that builds fast, incremental, error-tolerant syntax trees for source code; cyber tools use it to parse and query code without writing a full compiler frontend. TQL grammar adds bracket pair and verbatim indentation queries for editor integrations, and distinct highlight captures so editor themes can color pipeline operators, times, durations, IP addresses, and subnets apart from function calls and other literals.
Versioned packages with live rollout and configuration managementnew45how completely this was documenteddepth20/40specificity15/30actionability10/3045 / 100
Tenzir v7 supports versioned packages with live rollout and configuration management for deployments.
Tenzir adds ocsf_compact for lossless OCSFOCSFAn open schema standard, maintained by the OCSF community under CISA and industry backing, that defines a common event format so cyber tools can share and ingest log data without per-source normalization work. compaction, category_uid/type_uid computation in ocsf_derive, a version function, time-zone support in parse_time, automatic syslog framing detection, and lower memory and CPUCPUThe main general-purpose processor in a computer. It runs application instructions and coordinates work with memory and other hardware. usage.
Add lossless OCSF compaction with the ocsf_compact operator to reduce event size without losing fields.
Compute category_uid and type_uid automatically in the ocsf_derive operator.
Support time zones in the parse_time function.
Add a version function to retrieve the current Tenzir version inside a pipeline.
Automatic syslog framing detection added to the syslog parser.
+1 moreshow less
Lower base memory usage for pipelines and lower idle CPU usage for nodes.
6 details in Everything
└──▷ BREAKING ON UPGRADE
!
Several OCSF operators are deprecated in this release — existing pipelines using the deprecated forms may need updating.Our best guess
Tenzir adds a new heterogeneous data model for faster pipelines, typed AI decisions via ai_decide, secrets in function arguments, Parquet encryption, and new built-in functions including type_kind, Luhn, and CRC-32.
Enable typed decisions in pipelines with the new ai_decide operator, which now supports Decisions endpoints and typed output.
Pass secrets as function arguments and use secret keys with the hmac function, with secrets automatically censored on assignment.
Compute Luhn check digits and CRC-32 checksums with new built-in functions.
Encrypt Parquet output with a shared AES key.
Inspect the kind of a type at runtime with the new type_kind function.
+8 moreshow less
Tenzir's new data model lets lists hold values of different types (heterogeneous lists) and makes pipelines faster overall.
Use parallel load balancing across pipeline branches.
Secret seeds for Crypto-PAn pseudonymisation are now supported.
Encryption functions are now available for use in pipelines.
Networking and admin tools are now included in the container images.
Tines is an apps, agents and automations, in the complete environment for enterprise AI.
Tines added per-product-area access controls, letting teams on the same workspace hold different permissions across product areas.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
└──▷ WHAT SHIPPED ·
1 FEATURE0 OF 1 FEATUREmost completely described firstwhat's the number?
Each feature carries 0–100 for how completely the vendor documented
it — not how big or important the work is. A major capability
described in eight words scores low, and that is the finding.
depth0–40what it does and how it works or what changed
specificity0–30names real surfaces — APIs, flags, formats, limits, numbers
actionability0–30enough to go use it — a named endpoint, flag, or config key tops this; a UI path is a starting point
Code and config rank first by construction: a documented endpoint,
flag or config key scores at the top of specificity and actionability,
so it sorts above a feature described only in prose. Under
60 the notes go thin — everything below the “thinner
coverage” line is thinner documentation, not smaller work.
Hover any meter for that feature's three sub-scores.
0100
Per-product-area access permissionsnew33how completely this was documenteddepth15/40specificity8/30actionability10/3033 / 100
Tines now lets admins control access to each product area independently, so analysts and builders who share a team can hold different product permissions rather than a single uniform access level.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
Databricks previewed Power BI/Tableau import via Genie Code for compliance workspaces and introduced fine-grained Unity Catalog permissions for pipeline flows.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
Databricks is a unified data analytics platform that enables collaborative data engineering, machine learning, and analytics using Apache Spark in the cloud.
Databricks previewed Power BI/Tableau import via Genie Code for compliance workspaces and introduced fine-grained Unity Catalog permissions for pipeline flows.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
└──▷ WHAT SHIPPED ·
2 FEATURES0 OF 2 FEATURESmost completely described firstwhat's the number?
Each feature carries 0–100 for how completely the vendor documented
it — not how big or important the work is. A major capability
described in eight words scores low, and that is the finding.
depth0–40what it does and how it works or what changed
specificity0–30names real surfaces — APIs, flags, formats, limits, numbers
actionability0–30enough to go use it — a named endpoint, flag, or config key tops this; a UI path is a starting point
Code and config rank first by construction: a documented endpoint,
flag or config key scores at the top of specificity and actionability,
so it sorts above a feature described only in prose. Under
60 the notes go thin — everything below the “thinner
coverage” line is thinner documentation, not smaller work.
Hover any meter for that feature's three sub-scores.
0100
Flows as securable objects in Unity Catalognew85how completely this was documenteddepth25/40specificity30/30actionability30/3085 / 100
Pipeline flows can now be individually secured in Unity Catalog (Beta) using fine-grained privileges: READ FLOW, REFRESH FLOW, WRITE FLOW, APPLY TAG, and MANAGE on flows, or CREATE FLOW on schemas.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
Restrict a service principal to only triggering a specific pipeline flow refresh without granting broader pipeline access.
$ GRANT REFRESH FLOW ON FLOW my_catalog.my_schema.my_flow TO `service-principal@example.com`;
BI file import via Genie Codenew65how completely this was documenteddepth25/40specificity20/30actionability20/3065 / 100
Power BI and Tableau files can be imported as AI/BI dashboards and metric views using Genie Code. It is available now for workspaces with the compliance security profile enabled by turning on Import BI files using Genie Code under Previews in workspace settings, and will be on by default starting December 15, 2026.
Enable Genie Code BI file import early for a compliance security profile workspace before the December 15, 2026 default rollout.
📍In the workspace settings, go to Previews and enable Import BI files using Genie Code.
IT-Harvest launched the CX-Index, a market-cap-weighted stock index tracking 23 cybersecurity vendors.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
Data-driven analyst firm covering the entire cybersecurity industry, with 4,000+ vendors and 11,000+ products in one living research platform (the IT-Harvest Dashboard).
IT-Harvest launched the CX-Index, a market-cap-weighted stock index tracking 23 cybersecurity vendors.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
└──▷ WHAT SHIPPED ·
1 FEATURE1 OF 1 FEATUREmost completely described firstwhat's the number?
Each feature carries 0–100 for how completely the vendor documented
it — not how big or important the work is. A major capability
described in eight words scores low, and that is the finding.
depth0–40what it does and how it works or what changed
specificity0–30names real surfaces — APIs, flags, formats, limits, numbers
actionability0–30enough to go use it — a named endpoint, flag, or config key tops this; a UI path is a starting point
Code and config rank first by construction: a documented endpoint,
flag or config key scores at the top of specificity and actionability,
so it sorts above a feature described only in prose. Under
60 the notes go thin — everything below the “thinner
coverage” line is thinner documentation, not smaller work.
Hover any meter for that feature's three sub-scores.
0101
CX-Index stock index for cybersecurity vendorsnew70how completely this was documenteddepth25/40specificity25/30actionability20/3070 / 100
IT-Harvest added the CX-Index, a market-cap-weighted stock index tracking 23 cybersecurity vendors, modeled after the S&P 500 weighting methodology.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄ It is viewable at it-harvest.com/cx-index.
Defectdojo is an open-source Unified Vulnerability Management, DevSecOps & ASPM
DefectDojo 3.5.103 adds an Early Light parser; performance improvements speed up reimport JIRA lookups, tag indexing, and audit log history builds.
└──▷ GET THIS VERSION
$ git clone --branch 3.5.103 https://github.com/DefectDojo/django-DefectDojo.git
# already have the repo? check out this version:$ git checkout 3.5.103
New integrationDefectDojo adds an Early Light parser to import Early Light scan results via the standard importer pipeline.
Reimport now batches finding-group JIRA lookups in process_groups_for_all_findings to reduce round-trips.
Tag tables now index UPPER(name) so adding a tag no longer requires a full table scan.
Audit log history index builds now receive dedicated memory and workers to complete faster.
gvm-libs is a C library providing core functions and utilities for the Greenbone Vulnerability Management (GVM) framework and its components.
gvm-libs v23.16.0 adds programmatic control over OSP connection timeouts and retries, and significantly expands Active DirectoryActive DirectoryMicrosoft's directory service for Windows networks that centralises authentication and authorisation; cyber tools target or integrate with it because it controls identity and access across most enterprise environments. connector search capabilities.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
└──▷ WHAT SHIPPED ·
2 FEATURES0 OF 2 FEATURESmost completely described firstwhat's the number?
Each feature carries 0–100 for how completely the vendor documented
it — not how big or important the work is. A major capability
described in eight words scores low, and that is the finding.
depth0–40what it does and how it works or what changed
specificity0–30names real surfaces — APIs, flags, formats, limits, numbers
actionability0–30enough to go use it — a named endpoint, flag, or config key tops this; a UI path is a starting point
Code and config rank first by construction: a documented endpoint,
flag or config key scores at the top of specificity and actionability,
so it sorts above a feature described only in prose. Under
60 the notes go thin — everything below the “thinner
coverage” line is thinner documentation, not smaller work.
Hover any meter for that feature's three sub-scores.
0100
Active Directory connector search improvementsnew55how completely this was documenteddepth25/40specificity20/30actionability10/3055 / 100
The Active Directory connector now supports searching container objects, extending the scope of AD queries beyond flat group and user lookups; it also supports group DN membership searches, and enforces base DN confinement checks to restrict the scope of AD searches.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
— Names three distinct AD search behaviors but no API or config names.v23.16.0
0200
OSP connection timeout and retry controlsnew45how completely this was documenteddepth20/40specificity15/30actionability10/3045 / 100
New getter and setter functions were added to the osp library to programmatically get and set the OSP connection timeout, and to get and set the maximum number of OSP connection retries.
— Describes the capability but no exact function names given.v23.16.0
oxo now captures developer fields in store listing fingerprints, extending the data collected during app-store asset scans.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
└──▷ WHAT SHIPPED ·
1 FEATURE0 OF 1 FEATUREmost completely described firstwhat's the number?
Each feature carries 0–100 for how completely the vendor documented
it — not how big or important the work is. A major capability
described in eight words scores low, and that is the finding.
depth0–40what it does and how it works or what changed
specificity0–30names real surfaces — APIs, flags, formats, limits, numbers
actionability0–30enough to go use it — a named endpoint, flag, or config key tops this; a UI path is a starting point
Code and config rank first by construction: a documented endpoint,
flag or config key scores at the top of specificity and actionability,
so it sorts above a feature described only in prose. Under
60 the notes go thin — everything below the “thinner
coverage” line is thinner documentation, not smaller work.
Hover any meter for that feature's three sub-scores.
0100
Developer fields in store listing fingerprintsimproved30how completely this was documenteddepth15/40specificity10/30actionability5/3030 / 100
oxo now captures developer fields in store listing fingerprints, extending the data collected during app-store asset scans.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
— Names the field addition but no schema, scope or usage detail.v2.23.0
N-able N-central shipped a preview of multi-level patch policy management with automatic inheritance across organisational tiers, alongside a reworked Tags management UX with scope filtering and a unified creation flow.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
N-able N-central is a secure endpoints and automate IT operations at scale. N-central delivers unified endpoint management with advanced RMM capabilities for complex environments.
N-able N-central shipped a preview of multi-level patch policy management with automatic inheritance across organisational tiers, alongside a reworked Tags management UX with scope filtering and a unified creation flow.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
└──▷ WHAT SHIPPED ·
2 FEATURES0 OF 2 FEATURESmost completely described firstwhat's the number?
Each feature carries 0–100 for how completely the vendor documented
it — not how big or important the work is. A major capability
described in eight words scores low, and that is the finding.
depth0–40what it does and how it works or what changed
specificity0–30names real surfaces — APIs, flags, formats, limits, numbers
actionability0–30enough to go use it — a named endpoint, flag, or config key tops this; a UI path is a starting point
Code and config rank first by construction: a documented endpoint,
flag or config key scores at the top of specificity and actionability,
so it sorts above a feature described only in prose. Under
60 the notes go thin — everything below the “thinner
coverage” line is thinner documentation, not smaller work.
Hover any meter for that feature's three sub-scores.
0100
Multi-level patch policy management (preview)new84how completely this was documenteddepth35/40specificity24/30actionability25/3084 / 100
Patch policies now cascade automatically down the organisational hierarchy — Service Organisation, Customer, and Site — with more specific policies at lower levels overriding inherited ones when assets conflict.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄ The policy summary page in policy details shows exactly which assets a policy is deployed to before you deploy an override, and a new Add tags button on the Assets targeting step lets you create tags on the fly during policy creation without leaving the workflow.
Override a Service Organisation baseline policy for a customer or site with unique patching requirements, without affecting the rest of your infrastructure.
📍In the Policies grid, use the Scope Selector to switch from Service Organisation to the target Customer or Site, then create a new policy — it will automatically override the inherited SO-level policy only for assets at that scope.
Confirm which assets will be affected before activating a new policy override at the Customer or Site level.
📍Open the policy, go to its policy summary page in policy details, and review the effective assets list to verify deployment scope before saving.
Scope Picker and unified creation flow in Tags managementimproved68how completely this was documenteddepth28/40specificity20/30actionability20/3068 / 100
The Tags grid now has a Scope Picker at the top that filters tags by organisation, showing tags inherited from parent levels and those available to child scopes. Tag creation and assignment are combined into one flow — choosing type, scope (from the org tree), and assets all within the creation form with no context switching — and the grid gains an Assets column showing total asset counts per tag, including assets outside the user's access scope.
When managing tags across a multi-tenant hierarchy, use the Scope Picker to narrow the Tags grid to your organisation before creating or assigning a tag.
📍In N-central, go to Tags › Tags grid › Scope Picker (top of grid), select your organisation, then click Create Tag — set the type, scope, and asset assignment all within the single creation form.
Bbot is the recursive internet scanner for hackers.
BBOT 3.1 replaces massdns with a new BlastDNS resolution engine for roughly 40% faster subdomain scans, turns CDNCDNA content delivery network, servers distributed across locations that cache and serve content near readers. It can reduce loading time and traffic to the original server./WAFWAFA reverse proxy that inspects HTTP/HTTPS traffic and blocks requests matching known attack patterns such as SQL injection or XSS, sitting between clients and a web application to filter malicious input before it reaches the server. port filtering on by default, and ships seven new modules alongside a refreshed nuclei integration — plus a string of breaking changes including a Python 3.11 requirement and a reworked preset system.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
└──▷ WHAT SHIPPED ·
13 FEATURES2 OF 13 FEATURESmost completely described firstwhat's the number?
Each feature carries 0–100 for how completely the vendor documented
it — not how big or important the work is. A major capability
described in eight words scores low, and that is the finding.
depth0–40what it does and how it works or what changed
specificity0–30names real surfaces — APIs, flags, formats, limits, numbers
actionability0–30enough to go use it — a named endpoint, flag, or config key tops this; a UI path is a starting point
Code and config rank first by construction: a documented endpoint,
flag or config key scores at the top of specificity and actionability,
so it sorts above a feature described only in prose. Under
60 the notes go thin — everything below the “thinner
coverage” line is thinner documentation, not smaller work.
Hover any meter for that feature's three sub-scores.
0101
BlastDNS-powered DNS resolution and resolver controlnew93how completely this was documenteddepth35/40specificity28/30actionability30/3093 / 100
BBOT now uses BlastDNS for all DNS traffic including brute-forcing, replacing massdns; adaptive rate control cuts failure rates and finishes subdomain scans about 40% faster, limiting the resolver pool to 600 randomly chosen public resolvers per scan.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄ The new --custom-nameservers flag lets you pin BBOT to specific resolvers for normal DNS resolution instead of the default public pool.
Pin BBOT to your internal resolvers for normal DNS resolution instead of using the default public pool — useful in lab or air-gapped environments.
— Names mechanism, flag, and concrete 40%/600-resolver figures with runnable example.launch-20261112-a0e1a802
0202
Built-in CDN/WAF port filtering, replacing portfilter modulebreaking88how completely this was documenteddepth30/40specificity28/30actionability30/3088 / 100
CDN/WAF port filtering is now built into BBOT and on by default, dropping non-80/443 ports on cdn- or waf-tagged hosts; revert with the new --allow-cdn-ports flag. The standalone portfilter module is removed, so it must be dropped from existing presets.
js_unpacker module for obfuscated JavaScriptnew88how completely this was documenteddepth30/40specificity28/30actionability30/3088 / 100
New js_unpacker module transparently unpacks source maps, webpack, obfuscator.io, Next.js, and Dean Edwards' packer JavaScript, feeding results back into BBOT for more endpoint and secrets discovery; it has been added to the js-audit preset.
Run a full JS audit against a target to unpack obfuscated bundles and surface leaked secrets with TruffleHog — js_unpacker is now included in the js-audit preset.
Nuclei integration: new presets, pass-through options, cached indexingimproved70how completely this was documenteddepth25/40specificity25/30actionability20/3070 / 100
Adds nine new nuclei pass-through options and two new presets: nuclei-recommended (built on ProjectDiscovery's curated recommended profile) and nuclei-new. Nuclei template indexing is now cached, cutting startup time from about 35 seconds to near-instant.
— Names presets and startup-time improvement, but the nine options are unnamed.launch-20261112-a0e1a802
0500
Preset system: module_dirs and preset condition changesbreaking70how completely this was documenteddepth25/40specificity25/30actionability20/3070 / 100
module_dirs now only works as a top-level preset key, and BBOT prompts before loading each custom module folder (use -y to accept). Preset conditions are now read-only — they can warn or abort but can no longer modify the preset.
New passive subdomain enumeration modules: otilabs and scanmalwarenew60how completely this was documenteddepth20/40specificity25/30actionability15/3060 / 100
Adds otilabs (queries OTI Labs, requires an API key) and scanmalware (draws from ScanMalware's public scan archive) as new passive subdomain enumeration modules.
nowafpls module for WAF-bypass padding detectionnew50how completely this was documenteddepth20/40specificity20/30actionability10/3050 / 100
New nowafpls module detects when the WAF-bypass padding technique is usable and is integrated into lightfuzz and other modules to test for vulnerabilities through WAFs.
Technology naming and event tagging changesbreaking40how completely this was documenteddepth20/40specificity15/30actionability5/3040 / 100
CPECPECommon Platform Enumeration, an NIST-maintained naming scheme that gives software, hardware, and operating systems a structured, machine-readable identifier so cyber tools can match vulnerabilities to affected products consistently.-based technologies are now named by product, and in-scope events are no longer tagged affiliate.
The `portfilter` module is removed; CDN/WAF port filtering is now built in and on by default — remove `portfilter` from your presets and use `--allow-cdn-ports` to restore the old behavior.Our best guess
Aembit added embedded user authorization for MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface. servers via OAuth PKCEPKCEProof Key for Code Exchange, an OAuth safeguard that binds an authorization code to the client that requested it. An intercepted code cannot be redeemed without the client's matching verifier., new PrometheusPrometheusAn open-source monitoring system maintained by the CNCF that scrapes and stores time-series metrics, giving cyber tools a standard way to expose and query operational data. metrics covering MCP gateway credential injection and workload event batching, and clearer OIDCOIDCOpenID Connect, a sign-in protocol built on OAuth 2.0. An application uses identity information from a trusted provider instead of maintaining a separate password for each user. client ID visibility.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
Aembit is a runtime IAM for AI agents and workloads. Enforce access policies, keep downstream credentials out of agents, audit access, and revoke it with a kill switch.
Aembit added embedded user authorization for MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface. servers via OAuth PKCEPKCEProof Key for Code Exchange, an OAuth safeguard that binds an authorization code to the client that requested it. An intercepted code cannot be redeemed without the client's matching verifier., new PrometheusPrometheusAn open-source monitoring system maintained by the CNCF that scrapes and stores time-series metrics, giving cyber tools a standard way to expose and query operational data. metrics covering MCP gateway credential injection and workload event batching, and clearer OIDCOIDCOpenID Connect, a sign-in protocol built on OAuth 2.0. An application uses identity information from a trusted provider instead of maintaining a separate password for each user. client ID visibility.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
└──▷ WHAT SHIPPED ·
3 FEATURES1 OF 3 FEATURESmost completely described firstwhat's the number?
Each feature carries 0–100 for how completely the vendor documented
it — not how big or important the work is. A major capability
described in eight words scores low, and that is the finding.
depth0–40what it does and how it works or what changed
specificity0–30names real surfaces — APIs, flags, formats, limits, numbers
actionability0–30enough to go use it — a named endpoint, flag, or config key tops this; a UI path is a starting point
Code and config rank first by construction: a documented endpoint,
flag or config key scores at the top of specificity and actionability,
so it sorts above a feature described only in prose. Under
60 the notes go thin — everything below the “thinner
coverage” line is thinner documentation, not smaller work.
Hover any meter for that feature's three sub-scores.
0100
New Prometheus metrics for MCP gateway and event batchingnew85how completely this was documenteddepth30/40specificity30/30actionability25/3085 / 100
aembit_mcp_gateway_upstream_mcp_server_credential_injections_total counts requests to MCP servers by whether MCP Identity Gateway added the credential the Access Policy configures; aembit_mcp_gateway_upstream_mcp_server_http_requests_total now reports outcome="auth_error" when an MCP server returns an HTTP authentication error; aembit_event_batch_size is a new histogram recording how many workload events each batch carried for every batch that reached Aembit Cloud.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄ Batch size on a self-hosted MCP Identity Gateway can be capped with AEMBIT_EVENT_MAX_SEND_BATCH, e.g. AEMBIT_EVENT_MAX_SEND_BATCH=100.
Cap workload event batch size on a self-hosted MCP Identity Gateway to reduce per-flush payload size in high-traffic environments.
User authorization into MCP servers via OAuth PKCEnew75how completely this was documenteddepth30/40specificity25/30actionability20/3075 / 100
Authorize a person into a specific MCP server by naming it in the mcp_server_url parameter of a standard OAuth authorization code flow with PKCE — Aembit shows no login page, Identity Provider picker, or consent screen of its own.
Ping Identity's rapid-channel update adds named push profiles with new v2.0 journey nodes for multi-brand MFAMFAMulti-factor authentication, requiring proof from different categories, such as a password and a hardware key, to sign in. It reduces reliance on a password alone., a new control to disable may_act enforcement for token exchangetoken exchangeAn OAuth 2.0 protocol extension (RFC 8693) that lets a service trade one token for another with different scopes, audience, or type, enabling delegation and impersonation across service boundaries without re-authenticating the user., and SAMLSAMLA standard for passing signed identity assertions between a sign-in provider and an application. It lets a user sign in through their organization's identity provider.AuthnContextAuthnContextA SAML standard element that describes how a user was authenticated — for example, via password, smartcard, or MFA — letting a relying party enforce minimum assurance requirements before granting access. Class Ref prefixing control, alongside smaller performance and usability fixes to journey nodes and the registration flow.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
Ping Identity is an identity and access management platform providing authentication, single sign-on, multi-factor authentication, and identity governance services.
Ping Identity's rapid-channel update adds named push profiles with new v2.0 journey nodes for multi-brand MFAMFAMulti-factor authentication, requiring proof from different categories, such as a password and a hardware key, to sign in. It reduces reliance on a password alone., a new control to disable may_act enforcement for token exchangetoken exchangeAn OAuth 2.0 protocol extension (RFC 8693) that lets a service trade one token for another with different scopes, audience, or type, enabling delegation and impersonation across service boundaries without re-authenticating the user., and SAMLSAMLA standard for passing signed identity assertions between a sign-in provider and an application. It lets a user sign in through their organization's identity provider.AuthnContextAuthnContextA SAML standard element that describes how a user was authenticated — for example, via password, smartcard, or MFA — letting a relying party enforce minimum assurance requirements before granting access. Class Ref prefixing control, alongside smaller performance and usability fixes to journey nodes and the registration flow.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
└──▷ WHAT SHIPPED ·
7 FEATURES0 OF 7 FEATURESmost completely described firstwhat's the number?
Each feature carries 0–100 for how completely the vendor documented
it — not how big or important the work is. A major capability
described in eight words scores low, and that is the finding.
depth0–40what it does and how it works or what changed
specificity0–30names real surfaces — APIs, flags, formats, limits, numbers
actionability0–30enough to go use it — a named endpoint, flag, or config key tops this; a UI path is a starting point
Code and config rank first by construction: a documented endpoint,
flag or config key scores at the top of specificity and actionability,
so it sorts above a feature described only in prose. Under
60 the notes go thin — everything below the “thinner
coverage” line is thinner documentation, not smaller work.
Hover any meter for that feature's three sub-scores.
0100
Named push profiles for multi-brand MFA journeysnew80how completely this was documenteddepth35/40specificity30/30actionability15/3080 / 100
The Push Notification service now supports named push profiles, letting multiple brands or mobile applications share a single realm.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄ New v2.0 Push Sender, Push Registration, Combined MFA Registration, and Recovery Code Collector Decision journey nodes let you select which push profile a journey uses, with a Legacy Configuration profile available to preserve previous realm-level settings during migration.
When running multiple branded mobile apps in one realm, create a named push profile for each brand so journeys can target the correct push sender.
📍In the admin console, go to Realms › <your-realm> › Services › Push Notification, create a new named push profile, then add a v2.0 Push Sender node to your journey and select that profile.
Disable may_act enforcement for token exchangenew55how completely this was documenteddepth25/40specificity15/30actionability15/3055 / 100
A new control lets you disable may_act enforcement for token exchange at the realm or client level, allowing subject tokens from third-party identity providersidentity providersA service that authenticates users and supplies identity information to other applications. It is the sign-in authority behind many single sign-on systems. or authorization servers to be accepted.
SAML AuthnContext Class Ref prefixing controlnew55how completely this was documenteddepth20/40specificity20/30actionability15/3055 / 100
A new Prefix AuthnContext Class Ref configuration option on hosted SAML service providers controls whether custom AuthnContext Class Ref values are automatically prefixed, so SPs can send non-URN namespaced references unchanged to the IdPIdPA service that authenticates users and supplies identity information to other applications. It is the sign-in authority behind many single sign-on systems..
When a custom SAML AuthnContext Class Ref already uses a non-URN namespace, disable automatic URN prefixing so the reference reaches the IdP unchanged.
📍In the admin console, go to Realms › <your-realm> › Applications › <hosted-SP> › SAML 2.0, open Advanced settings, and disable Prefix AuthnContext Class Ref.
Reduced data requests in Identify Existing User and Query Filter Decision nodesimproved50how completely this was documenteddepth25/40specificity20/30actionability5/3050 / 100
Version 2.0 of the Identify Existing User node and version 2.0 of the Query Filter Decision node each reduce the data requested (to identify a user and to evaluate a query filter, respectively), improving journey performance.
Hyphenated identifiers in PingOne product connectionsimproved40how completely this was documenteddepth25/40specificity10/30actionability5/3040 / 100
PingOne product connections now accept identifiers containing hyphens, so hyphenated PingOne environment names can be connected and deleted without validation errors.
Visual password requirement indicators in registration journeyimproved30how completely this was documenteddepth20/40specificity10/30actionability0/3030 / 100
The default registration journey now shows password requirement status using checkmarks and Xs rather than color alone.
1Password: Manage 1Password Environments from the CLI with the new op environment command, supporting create, get, list, edit, and delete operations.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
└──▷ TRY IT
List all 1Password Environments in a vault to audit what environments exist across your org.
$ op environment list
Retrieve a specific secret from a 1Password Environment for use in a CI pipeline or script.
$ op environment secret get <secret-name> --environment <environment-name>
JumpCloud makes it simple to manage Windows, Apple, Linux, and Android devices and to deliver secure access with SSO, MFA, Zero Trust, RADIUS, and more.
JumpCloud's November 2026 release adds audit-ready Access CertificationAccess CertificationA periodic review process in which managers or system owners confirm or revoke which users should retain access to systems and data, reducing privilege creep over time. campaigns, an AI-agent delegated access model, SIEMSIEMSecurity information and event management, collecting logs from many systems and correlating events to find suspicious activity. Analysts use it to investigate alerts and incidents.-ready Directory Insights log export, and user-level Windows policy controls, alongside incremental upgrades to Workflows, Device Lock, App Catalog, SCIMSCIMA standard for creating, updating and removing user accounts and groups across applications. It helps an identity provider keep access in sync when people join, change roles or leave. provisioning and browser discovery.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
└──▷ WHAT SHIPPED ·
12 FEATURES3 OF 12 FEATURESmost completely described firstwhat's the number?
Each feature carries 0–100 for how completely the vendor documented
it — not how big or important the work is. A major capability
described in eight words scores low, and that is the finding.
depth0–40what it does and how it works or what changed
specificity0–30names real surfaces — APIs, flags, formats, limits, numbers
actionability0–30enough to go use it — a named endpoint, flag, or config key tops this; a UI path is a starting point
Code and config rank first by construction: a documented endpoint,
flag or config key scores at the top of specificity and actionability,
so it sorts above a feature described only in prose. Under
60 the notes go thin — everything below the “thinner
coverage” line is thinner documentation, not smaller work.
Hover any meter for that feature's three sub-scores.
0101
Access Certification campaigns with audit-ready exportsnew93how completely this was documenteddepth35/40specificity28/30actionability30/3093 / 100
JumpCloud adds an Access Certification feature supporting one-time or recurring review campaigns for SSOSSOSingle sign-on, using one identity provider to sign in to multiple applications. It reduces separate passwords and centralizes account and access management. applications and directory user groups, multi-stage approval workflows, automated access revocation or identity suspension, and a centralized admin dashboard tracking pending reviews.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄ Completed campaigns can be exported as tamper-proof snapshots in CSV, JSON, or PDF formats with SHA-256SHA-256A cryptographic hash function that turns data into a fixed 256-bit fingerprint. It is used to check integrity; hashing does not encrypt the original data. document hashing and an administrator sign-off flow for compliance audits.
Export a frozen, audit-ready Access Certification campaign snapshot in PDF with SHA-256 hashing for a compliance audit.
📍In the Admin Portal, go to Access Certification › open the completed campaign › select Export › choose PDF format › complete the administrator sign-off flow to download the tamper-proof snapshot.
Agent Delegated Access for AI agentsnew85how completely this was documenteddepth30/40specificity25/30actionability30/3085 / 100
JumpCloud adds Agent Delegated Access (Private Preview), letting an AI agent authenticate with its own JumpCloud-managed identity while acting on behalf of an end user, using token exchangetoken exchangeAn OAuth 2.0 protocol extension (RFC 8693) that lets a service trade one token for another with different scopes, audience, or type, enabling delegation and impersonation across service boundaries without re-authenticating the user. and allowed-audiences configuration, with direct API access without routing calls through AI Gateway.
Enable Agent Delegated Access for an AI agent so it can act on behalf of a user — configure a trusted OIDCOIDCOpenID Connect, a sign-in protocol built on OAuth 2.0. An application uses identity information from a trusted provider instead of maintaining a separate password for each user. application and allowed audiences from the agent's Delegated access tab in the Admin Portal.
📍In the Admin Portal, open the agent application › select the Delegated access tab › enable delegation › select a trusted OIDC application › configure Allowed audiences for the resources the agent may request tokens for › save.
SIEM integration for Directory Insightsnew73how completely this was documenteddepth30/40specificity28/30actionability15/3073 / 100
JumpCloud adds SIEM integration in Directory Insights for continuous identity-event ingestion using an existing API key, supporting RAW and OCSFOCSFAn open schema standard, maintained by the OCSF community under CISA and industry backing, that defines a common event format so cyber tools can share and ingest log data without per-source normalization work.-normalized formats with JSON or CSV delivery, plus asynchronous historical log export jobs covering up to 7-day windows across a 13-month retention lookback with email notification on completion.
Device Lock enhancements for Apple MDMimproved70how completely this was documenteddepth25/40specificity25/30actionability20/3070 / 100
Device Lock for Apple MDMMDMMobile device management, centrally configuring and managing enrolled devices. Administrators use it to apply settings, distribute applications and enforce device security policies. now supports automatic PIN generation and escrow for macOS, retrievable from the Admin Portal under Highlights > Device Lock > View PIN, an optional Lock Screen message and phone number for Mac, iPhone, and iPad, and a macOS Lock Status column in the Devices list.
User-level Windows registry and MDM policiesnew65how completely this was documenteddepth25/40specificity25/30actionability15/3065 / 100
JumpCloud adds a User Level Windows Custom Registry Keys Policy that applies settings under HKEY_CURRENT_USER when an assigned user signs in, without requiring MDM enrollment or a device restart, and a User Level Custom Windows MDM (OMA-URI) Policy for configuring custom MDM settings scoped to a user's profile on MDM-enrolled Windows devices.
OAuth 2.0 authentication for Workflow connectorsnew65how completely this was documenteddepth25/40specificity25/30actionability15/3065 / 100
Workflow connectors now support OAuth 2.0OAuth 2.0An open authorization framework (IETF RFC 6749) that lets a service grant limited API access to a third party without sharing credentials, giving cyber tools a standard way to authenticate against external platforms. authentication, including Client CredentialsClient CredentialsAn OAuth 2.0 grant type where a service authenticates directly with an authorization server using its own ID and secret, with no user involved, to obtain an access token for machine-to-machine API calls. and Authorization Code with PKCEPKCEProof Key for Code Exchange, an OAuth safeguard that binds an authorization code to the client that requested it. An intercepted code cannot be redeemed without the client's matching verifier., with credentials stored in Secrets Manager and tokens refreshed automatically.
MCP-enabled AI client management of Workflowsnew60how completely this was documenteddepth25/40specificity20/30actionability15/3060 / 100
JumpCloud Workflows can now be created, modified, duplicated, and troubleshot through a connected MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface.-enabled AI client using natural language, including inspecting run history and manually triggering workflows.
Per-application group name override in SCIMnew55how completely this was documenteddepth20/40specificity20/30actionability15/3055 / 100
Admins can set a per-application group name override using Show In App As on the User Groups and Agent Groups tab in SCIM provisioning, without changing the underlying JumpCloud group name; the override takes priority over Group Export Attribute Mapping.
Expanded Workflows template librarynew55how completely this was documenteddepth20/40specificity20/30actionability15/3055 / 100
The Workflows template library now includes templates for auto-suspending users on inactivity, blocking newly discovered generative AI apps, revoking Mac admin privileges on enrollment, sensitive software install email alerts, user group membership emails, pre-hire holding group placement, and setting device custom attributes on agent enrollment.
Capability-based app setup for SSO, SCIM, MCPnew55how completely this was documenteddepth20/40specificity20/30actionability15/3055 / 100
Capability-based app setup now lets admins select SSO, SCIM, and/or MCP independently when adding an application, and create MCP-only applications available through AI Gateway without configuring SSO.
Browser discovery via JumpCloud Go extensionnew55how completely this was documenteddepth20/40specificity20/30actionability15/3055 / 100
Browser discovery in AI & SaaS ManagementSaaS ManagementA category of software that tracks, governs, and optimises an organisation's portfolio of SaaS subscriptions, covering spend, access, and licence utilisation from a single control plane. can now start from a JumpCloud-managed device sign-in or via the JumpCloud Go extension deployed through a third-party MDM, using a Client ID and Client secret from Settings.
App Catalog update notification and self-service controlsimproved55how completely this was documenteddepth20/40specificity20/30actionability15/3055 / 100
App Catalog now lets admins disable end-user update notifications per application with Disable end user update notifications, and lets users initiate required-app updates from the Self Service App Catalog.
Yubico makes YubiKey hardware security keys that provide strong two-factor and passwordless authentication to protect accounts and sensitive systems.
Yubico introduced two early-access offerings at YubiNation 2026: an identity-lifecycle trust solution called YubiID and a post-quantumpost-quantumCryptography designed to stay secure against a large quantum computer, which could break today's RSA and elliptic-curve keys. NIST standardized the first algorithms in 2024 as ML-KEM, ML-DSA and SLH-DSA.-ready YubiKey.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
└──▷ WHAT SHIPPED ·
2 FEATURES2 OF 2 FEATURESmost completely described firstwhat's the number?
Each feature carries 0–100 for how completely the vendor documented
it — not how big or important the work is. A major capability
described in eight words scores low, and that is the finding.
depth0–40what it does and how it works or what changed
specificity0–30names real surfaces — APIs, flags, formats, limits, numbers
actionability0–30enough to go use it — a named endpoint, flag, or config key tops this; a UI path is a starting point
Code and config rank first by construction: a documented endpoint,
flag or config key scores at the top of specificity and actionability,
so it sorts above a feature described only in prose. Under
60 the notes go thin — everything below the “thinner
coverage” line is thinner documentation, not smaller work.
Hover any meter for that feature's three sub-scores.
0101
PQC-Ready YubiKey public betanew35how completely this was documenteddepth15/40specificity12/30actionability8/3035 / 100
Yubico launched a PQC-Ready YubiKey (Public Beta), hardware engineered to operationalize post-quantum defenses without disrupting existing day-to-day workflows.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
YubiID identity-lifecycle trust solutionnew33how completely this was documenteddepth15/40specificity10/30actionability8/3033 / 100
Yubico added YubiID (Limited Early Access), a solution that carries verifiable trust across the complete identity lifecycle — from initial onboarding to account recovery.
Okta Identity Engine is a secure, scalable, and highly available authentication and user management for any app.
Okta Identity Engine shipped generally available passkeypasskeyPhishing-resistant sign-in credentials built on FIDO2/WebAuthn: the device holds a private key and proves possession instead of the user typing a password. They sync across a user's devices. (FIDO2 WebAuthnFIDO2 WebAuthnA W3C standard that lets browsers authenticate users with public-key cryptography via hardware tokens or platform authenticators, eliminating password transmission and making phishing credential theft structurally impossible.) support for user identification policies and expanded AI agent configuration across identification, platform mapping, and custom claims.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
└──▷ WHAT SHIPPED ·
2 FEATURES0 OF 2 FEATURESmost completely described firstwhat's the number?
Each feature carries 0–100 for how completely the vendor documented
it — not how big or important the work is. A major capability
described in eight words scores low, and that is the finding.
depth0–40what it does and how it works or what changed
specificity0–30names real surfaces — APIs, flags, formats, limits, numbers
actionability0–30enough to go use it — a named endpoint, flag, or config key tops this; a UI path is a starting point
Code and config rank first by construction: a documented endpoint,
flag or config key scores at the top of specificity and actionability,
so it sorts above a feature described only in prose. Under
60 the notes go thin — everything below the “thinner
coverage” line is thinner documentation, not smaller work.
Hover any meter for that feature's three sub-scores.
0100
Passkey support for user identification policiesnew40how completely this was documenteddepth15/40specificity15/30actionability10/3040 / 100
Passkey (FIDO2FIDO2The open standard behind passkeys and hardware security keys, where a website checks a signature from the user's device instead of a password. Because the key is bound to the site, phishing pages cannot use it.WebAuthnWebAuthnA web standard that lets a browser prove possession of a passkey or hardware key to a website. It is the browser-facing part of FIDO2 sign-in.) support for user identification policies is now generally available in Production.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
AI agent configuration expandedimproved38how completely this was documenteddepth15/40specificity15/30actionability8/3038 / 100
AI agents now support a source property for identification and configuration, multiple platform mappings with external ID filtering, and AI agent attributes are now available in custom claims (GA in preview).
Commvault Cloud expanded Protection Groups to cover Amazon Redshift, DynamoDB, and multi-account Azure/GCP onboarding, added SaaSSaaSSoftware delivered over the internet and billed by subscription, hosted and operated by the vendor so the buyer skips installation and infrastructure management. protection for ServiceNow and Zscaler, and introduced AWS European Sovereign Cloud support, Redshift Serverless backups, RDS for SQL Server transaction log backups, and Threat Scan findings triage.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
Commvault Cloud is a backup, recovery, and cyber-resilience platform for hybrid, multi-cloud, and SaaS workloads.
Commvault Cloud expanded Protection Groups to cover Amazon Redshift, DynamoDB, and multi-account Azure/GCP onboarding, added SaaSSaaSSoftware delivered over the internet and billed by subscription, hosted and operated by the vendor so the buyer skips installation and infrastructure management. protection for ServiceNow and Zscaler, and introduced AWS European Sovereign Cloud support, Redshift Serverless backups, RDS for SQL Server transaction log backups, and Threat Scan findings triage.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
└──▷ WHAT SHIPPED ·
8 FEATURES0 OF 8 FEATURESmost completely described firstwhat's the number?
Each feature carries 0–100 for how completely the vendor documented
it — not how big or important the work is. A major capability
described in eight words scores low, and that is the finding.
depth0–40what it does and how it works or what changed
specificity0–30names real surfaces — APIs, flags, formats, limits, numbers
actionability0–30enough to go use it — a named endpoint, flag, or config key tops this; a UI path is a starting point
Code and config rank first by construction: a documented endpoint,
flag or config key scores at the top of specificity and actionability,
so it sorts above a feature described only in prose. Under
60 the notes go thin — everything below the “thinner
coverage” line is thinner documentation, not smaller work.
Hover any meter for that feature's three sub-scores.
0100
Multi-account onboarding in Azure and GCP Protection Groupsnew75how completely this was documenteddepth30/40specificity30/30actionability15/3075 / 100
Azure Protection Groups and GCP Protection Groups now support onboarding multiple Azure subscriptions or GCP projects in a single workflow, with rule-based discovery and automatic protection for Azure Virtual Machines, Blob Storage, Azure Files, Azure Data Lake Storage (ADLS), CloudSQL, BigQueryBigQueryA serverless, cloud-hosted analytical database from Google that runs SQL queries across massive datasets without managing infrastructure, giving cyber tools scalable storage and fast querying over large volumes of log or event data., Bigtable, Spanner, and AlloyDB.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
— Enumerates every supported service but no exact setup steps2026-11-13-ffc0a6fb
0200
Amazon Redshift Serverless snapshot and export backupsnew75how completely this was documenteddepth30/40specificity30/30actionability15/3075 / 100
Amazon Redshift Serverless data and metadata can now be protected with snapshot and export-based backups, with snapshots supporting cross-region recovery and exported data stored in Commvault AirGap on Amazon S3.
— Names backup types, storage target and recovery mechanism2026-11-13-ffc0a6fb
0300
SaaS protection for ServiceNow and Zscalernew60how completely this was documenteddepth25/40specificity25/30actionability10/3060 / 100
Commvault now protects and recovers ServiceNow apps and records for organizations that rely on ServiceNow for IT service management, and protects Zscaler security policies, access controls, and configurations, enabling recovery from accidental changes or configuration errors.
— Names both platforms and protected objects but no setup steps2026-11-13-ffc0a6fb
thinner coverage below
0400
Redshift and DynamoDB support in AWS Protection Groupsnew55how completely this was documenteddepth20/40specificity20/30actionability15/3055 / 100
AWS Protection Groups now cover Amazon Redshift and DynamoDB, letting you onboard and protect analytics and application databases across multiple AWS accounts with rule-based policies.
— Names services and mechanism but no exact UI path or command2026-11-13-ffc0a6fb
0500
Cloud Application Recovery native workflows with expanded Azure coverageimproved55how completely this was documenteddepth25/40specificity20/30actionability10/3055 / 100
Cloud Application Recovery (formerly Cloud Rewind) now uses native Commvault protection and recovery workflows with 3x Azure resource coverage, including support for Azure Government.
— States scope and multiplier but no concrete mechanism detail2026-11-13-ffc0a6fb
0600
Transaction log backups for RDS for SQL Servernew55how completely this was documenteddepth20/40specificity20/30actionability15/3055 / 100
Commvault now supports transaction log (t-log) backups for Amazon RDS for SQL Server, enabling point-in-time database restore.
— Specific service and outcome but no exact command shown2026-11-13-ffc0a6fb
0700
Threat Scan findings triage and failed-scan visibilityimproved55how completely this was documenteddepth25/40specificity15/30actionability15/3055 / 100
Threat Scan now lets users mark individual or multiple findings as safe to remove non-actionable items from the dashboard while preserving original detections in the index, and surfaces files that failed to scan.
— Describes UI action and effect but no exact navigation path2026-11-13-ffc0a6fb
0800
AWS European Sovereign Cloud support for database workloadsnew45how completely this was documenteddepth15/40specificity20/30actionability10/3045 / 100
Amazon cloud database workloads can now be protected and recovered in AWS European Sovereign Cloud (ESC), supporting European data residency and sovereignty requirements.
Kasten K10 9.0.11 adds HTTP(S) proxy support for outbound connections with automatic OpenShift CA extraction, enriches the location profile picker, and tightens security around the multi-cluster and console plugin proxies with two breaking changes.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
Kasten K10 is a veeam Kasten is a Kubernetes-native platform for backing up, restoring, and migrating containerized applications and their persistent data.
Kasten K10 9.0.11 adds HTTP(S) proxy support for outbound connections with automatic OpenShift CA extraction, enriches the location profile picker, and tightens security around the multi-cluster and console plugin proxies with two breaking changes.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
└──▷ WHAT SHIPPED ·
4 FEATURES0 OF 4 FEATURESmost completely described firstwhat's the number?
Each feature carries 0–100 for how completely the vendor documented
it — not how big or important the work is. A major capability
described in eight words scores low, and that is the finding.
depth0–40what it does and how it works or what changed
specificity0–30names real surfaces — APIs, flags, formats, limits, numbers
actionability0–30enough to go use it — a named endpoint, flag, or config key tops this; a UI path is a starting point
Code and config rank first by construction: a documented endpoint,
flag or config key scores at the top of specificity and actionability,
so it sorts above a feature described only in prose. Under
60 the notes go thin — everything below the “thinner
coverage” line is thinner documentation, not smaller work.
Hover any meter for that feature's three sub-scores.
0100
HTTP(S) proxy support with OpenShift CA auto-extractionnew85how completely this was documenteddepth30/40specificity25/30actionability30/3085 / 100
Kasten K10 now supports HTTP(S) proxies for outbound connections.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄ On OpenShift, setting global.proxy.openshift.caCertsAutoExtraction=true automatically extracts the cluster-wide proxy's trustedCA bundle and cluster CA certificates into Kasten's trust bundle, so TLS-intercepting proxies work without manual certificate configuration.
Enable automatic CA extraction on OpenShift so a TLS-intercepting proxy is trusted without manually distributing certificates.
— Exact config key and value given, directly runnable9.0.11
0200
Multi-cluster proxy no longer forwards session credentialsbreaking60how completely this was documenteddepth30/40specificity20/30actionability10/3060 / 100
The multi-cluster proxy no longer forwards the caller's session and authentication credentials to secondary clusters, no longer relays cookie-changing response headers from them to the browser, and WebSocketWebSocketA protocol that upgrades an HTTP connection into a persistent, full-duplex channel, letting a server push data to a client without repeated polling. Cyber tools use it for real-time event streaming and live command output. and other protocol upgrades are no longer supported through the proxy.
— Explains behavior change but gives no migration step9.0.11
thinner coverage below
0300
OpenShift console plugin proxy drops plain-HTTP port 8080breaking50how completely this was documenteddepth20/40specificity20/30actionability10/3050 / 100
The OpenShift console plugin proxy's plain-HTTP listener on port 8080 is no longer reachable from the openshift-console namespace.
— Names exact port and namespace, no remediation given9.0.11
0400
Richer location profile selection dropdownimproved40how completely this was documenteddepth15/40specificity15/30actionability10/3040 / 100
The location profile selection dropdown now shows each profile's provider type, target, region, and immutability status inline.
— Describes UI change but no exact navigation path9.0.11
└──▷ BREAKING ON UPGRADE
!
The multi-cluster proxy no longer forwards the caller's session and authentication credentials to secondary clusters, no longer relays cookie-changing response headers from them to the browser, and WebSocket and other protocol upgrades are no longer supported through the proxy.Our best guess
Tailscale's Terraform Provider gained a new attribute for controlling tailnet route selection, including enabling MagicRoute.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
Tailscale creates private WireGuard-based mesh networks with centralized identity and access controls.
Tailscale's Terraform Provider gained a new attribute for controlling tailnet route selection, including enabling MagicRoute.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
└──▷ WHAT SHIPPED ·
1 FEATURE0 OF 1 FEATUREmost completely described firstwhat's the number?
Each feature carries 0–100 for how completely the vendor documented
it — not how big or important the work is. A major capability
described in eight words scores low, and that is the finding.
depth0–40what it does and how it works or what changed
specificity0–30names real surfaces — APIs, flags, formats, limits, numbers
actionability0–30enough to go use it — a named endpoint, flag, or config key tops this; a UI path is a starting point
Code and config rank first by construction: a documented endpoint,
flag or config key scores at the top of specificity and actionability,
so it sorts above a feature described only in prose. Under
60 the notes go thin — everything below the “thinner
coverage” line is thinner documentation, not smaller work.
Hover any meter for that feature's three sub-scores.
0100
Route selection control in Terraform providernew75how completely this was documenteddepth20/40specificity25/30actionability30/3075 / 100
The tailscale_tailnet_settings resource now supports a route_selection attribute to set the route selection method for a tailnet, including enabling MagicRoute, e.g.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄ route_selection = "magicroute".
Enable MagicRoute for your tailnet by setting the new route_selection attribute in Terraform.
Palo Alto Prisma Access is a cloud-delivered security platform providing secure remote access, network security, and zero-trust protection for users and applications globally.
Palo Alto Networks launched Sovereign Prisma SASESASEA network architecture model combining wide-area networking and network security functions into a single cloud-delivered service, letting cyber tools enforce consistent policy regardless of where users or workloads sit., a regulated-enterprise variant of Prisma SASE adding EU data residency, multi-cloud failover, on-premises processing options, customer-held encryption keys, and compliance-oriented audit logging.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
└──▷ WHAT SHIPPED ·
5 FEATURES5 OF 5 FEATURESmost completely described firstwhat's the number?
Each feature carries 0–100 for how completely the vendor documented
it — not how big or important the work is. A major capability
described in eight words scores low, and that is the finding.
depth0–40what it does and how it works or what changed
specificity0–30names real surfaces — APIs, flags, formats, limits, numbers
actionability0–30enough to go use it — a named endpoint, flag, or config key tops this; a UI path is a starting point
Code and config rank first by construction: a documented endpoint,
flag or config key scores at the top of specificity and actionability,
so it sorts above a feature described only in prose. Under
60 the notes go thin — everything below the “thinner
coverage” line is thinner documentation, not smaller work.
Hover any meter for that feature's three sub-scores.
0101
Hybrid SASE Private Location for on-prem processingnew55how completely this was documenteddepth25/40specificity20/30actionability10/3055 / 100
Sovereign Prisma SASE adds a hybrid architecture option called SASE Private Location, letting organisations process sensitive traffic on-premises and hold their own encryption keys while still feeding anonymised telemetry to global threat intelligence.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
EU data residency boundary for Sovereign SASEnew48how completely this was documenteddepth25/40specificity18/30actionability5/3048 / 100
Sovereign Prisma SASE establishes an EU-region data boundary that restricts administrative access to EU-only personnel and performs in-jurisdiction TLS break-and-inspect, so all traffic processing stays within European jurisdiction.
— Explains mechanism and scope but no config surface named2026-11-11-15987456
0303
Multi-cloud sovereign backbone with failovernew45how completely this was documenteddepth25/40specificity15/30actionability5/3045 / 100
Sovereign Prisma SASE adds a multi-cloud backbone that uses multiple sovereign cloud service providers within the same jurisdiction, so traffic automatically fails over to an alternate provider without leaving the data-residency boundary if one cloud goes down.
— Describes failover mechanism, no concrete configuration steps given2026-11-11-15987456
0404
Hold Your Own Key encryption modelnew43how completely this was documenteddepth20/40specificity18/30actionability5/3043 / 100
Sovereign Prisma SASE introduces a Hold Your Own Key encryption model giving organisations direct cryptographic control over their own keys, without hairpinning traffic through external HSMsHSMsA tamper-resistant device that generates and stores cryptographic keys and performs operations with them, so the private key never leaves the hardware. Cloud KMS services run on fleets of them..
CISO Assistant is a one-stop-shop GRC platform for Risk Management, AppSec, Compliance & Audit, TPRM, BIA, Privacy, and Reporting. It supports 200+ global frameworks with automatic control mapping, including ISO 27001, NIST CSF, SOC 2, CIS, PCI DSS, NIS2, DORA, GDPR, HIPAA, CMMC, and more.
CISO Assistant adds OIDCOIDCOpenID Connect, a sign-in protocol built on OAuth 2.0. An application uses identity information from a trusted provider instead of maintaining a separate password for each user.-federated service accounts, OAuth-secured MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface., an evidence-review MCP tool, TPRMTPRMAssessing and monitoring the security risk an organization takes on through its vendors, suppliers and service providers. It covers the questionnaires, evidence and continuous checks behind letting a third party hold your data. workspace task filing, and configurable framework defaults.
└──▷ GET THIS VERSION
$ git clone --branch v4.3.0 https://github.com/intuitem/ciso-assistant-community.git
# already have the repo? check out this version:$ git checkout v4.3.0
Add service accounts federated with an external OIDC identity provideridentity providerA service that authenticates users and supplies identity information to other applications. It is the sign-in authority behind many single sign-on systems. for non-interactive, machine-to-machine authentication.
Secure the MCP server with OAuth sign-in so that MCP clients must authenticate before accessing any tool.
Add an evidence-review tool to the MCP server, letting AI agents inspect and act on compliance evidence.
File tasks raised on a third-party audit directly in that third party's workspace via the TPRM module.
Surface extra settings for framework default behaviour, giving administrators control over per-framework defaults.
+3 moreshow less
Add new fields to two security exception notification templates.
Improve framework insights with richer analytics on the framework detail view.
Improve control plan and task occurrence handling.
OneTrust's Fall 2026 release adds a broad set of new Developer Portal APIs — covering DROP management for data brokers, document exceptions, mobile app consent, and role discovery — alongside enhancements to cookie scanning, consent templates, and script publishing.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
The OneTrust Developer Portal allows you to leverage APIs and SDKs to integrate external systems seamlessly and streamline the data flow with the OneTrust platform.
OneTrust's Fall 2026 release adds a broad set of new Developer Portal APIs — covering DROP management for data brokers, document exceptions, mobile app consent, and role discovery — alongside enhancements to cookie scanning, consent templates, and script publishing.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
└──▷ WHAT SHIPPED ·
8 FEATURES0 OF 8 FEATURESmost completely described firstwhat's the number?
Each feature carries 0–100 for how completely the vendor documented
it — not how big or important the work is. A major capability
described in eight words scores low, and that is the finding.
depth0–40what it does and how it works or what changed
specificity0–30names real surfaces — APIs, flags, formats, limits, numbers
actionability0–30enough to go use it — a named endpoint, flag, or config key tops this; a UI path is a starting point
Code and config rank first by construction: a documented endpoint,
flag or config key scores at the top of specificity and actionability,
so it sorts above a feature described only in prose. Under
60 the notes go thin — everything below the “thinner
coverage” line is thinner documentation, not smaller work.
Hover any meter for that feature's three sub-scores.
0100
Per-domain script file generation controlbreaking83how completely this was documenteddepth30/40specificity28/30actionability25/3083 / 100
Passing the new useDistinctScriptSrc request parameter in the Publish Script to Website API generates a unique otSDKStub.js per domain and changes the test script source URL.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄ This is a breaking change: any otSDKStub.js already implemented on a website must be updated to the new URL after enabling this parameter.
Mobile App Consent APIs with Web SDK supportnew78how completely this was documenteddepth28/40specificity30/30actionability20/3078 / 100
Six new Consent Management Platform APIs retrieve and manage mobile app consent experiences: Get Banner Data API, Get Preference Center Data API, Get IAB and Google Vendors API, Log Consent API, Get Universal Consent Purposes API, and Submit Global Opt-Out Request API. Web SDK support was added across four of these — Log Consent API, Get IAB and Google Vendors API, Get Banner Data API, and Get Preference Center Data API — to deliver consistent consent experiences across platforms.
Document exceptions search APIs in Tech Risk & Compliancenew75how completely this was documenteddepth25/40specificity28/30actionability22/3075 / 100
Search and retrieve document exceptions across Tech Risk & Compliance with the new Search Exceptions API and Get Exception API, now accessible via the EXCEPTION and EXCEPTION_READ OAuth scopes configured in Global Settings.
Cookie scan removal detection and IABC template typesimproved75how completely this was documenteddepth25/40specificity30/30actionability20/3075 / 100
The Get List of Added or Removed Cookies API now returns a new isNewlyRemoved response parameter to detect newly removed cookies in scan results. The Get Template Details API gains support for IABIABInitial Access Broker — a threat actor who specialises in compromising networks and selling that access to other criminals, such as ransomware operators, rather than exploiting it directly. Canada (IABC) display group types via six new type enum values: IABC_PURPOSE, IABC_STACK, IABC_FEATURE, IABC_SPL_PURPOSE, IABC_SPL_FEATURE, and IABC_DATA_CATEGORY.
DROP Management APIs for data brokersnew71how completely this was documenteddepth25/40specificity28/30actionability18/3071 / 100
Three new APIs let California-registered data brokers manage permanent deletion and opt-out requests: Create Data Subject Request for DROP Record API, Update DROP Record API, and Get List of DROP Records API.
Add options to selection-type data elementsnew53how completely this was documenteddepth20/40specificity18/30actionability15/3053 / 100
The new Add Data Element Option API lets you add new options to selection-type data elements without affecting existing data subjects or historical consent receipts.
Enabling `useDistinctScriptSrc` in the `Publish Script to Website API` changes the test script source URL — any `otSDKStub.js` already implemented on the website must be updated to the new URL.Our best guess
Compyl's latest release adds a new Coupa integration for supplier and approval-chain data, expands Netskope and AWS sync coverage, lets incidents link to multiple third parties, extends custom field search/filter to more grids, and locks down internal workflow-engine APIs behind authentication.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
Compyl is the AI-powered GRC platform that maps one control library to SOC 2, ISO 27001, HIPAA, PCI DSS and 70+ frameworks, so you stay audit-ready year-round.
Compyl's latest release adds a new Coupa integration for supplier and approval-chain data, expands Netskope and AWS sync coverage, lets incidents link to multiple third parties, extends custom field search/filter to more grids, and locks down internal workflow-engine APIs behind authentication.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
└──▷ WHAT SHIPPED ·
5 FEATURES0 OF 5 FEATURESmost completely described firstwhat's the number?
Each feature carries 0–100 for how completely the vendor documented
it — not how big or important the work is. A major capability
described in eight words scores low, and that is the finding.
depth0–40what it does and how it works or what changed
specificity0–30names real surfaces — APIs, flags, formats, limits, numbers
actionability0–30enough to go use it — a named endpoint, flag, or config key tops this; a UI path is a starting point
Code and config rank first by construction: a documented endpoint,
flag or config key scores at the top of specificity and actionability,
so it sorts above a feature described only in prose. Under
60 the notes go thin — everything below the “thinner
coverage” line is thinner documentation, not smaller work.
Hover any meter for that feature's three sub-scores.
0100
Coupa integration for supplier and approval datanew85how completely this was documenteddepth30/40specificity30/30actionability25/3085 / 100
A new Coupa integration syncs supplier requests, custom fields, contacts, approval chains, users, lookup values, and suppliers via OAuth, configured with your instance URL, client ID, client secret, and an optional form filter.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
Connect Compyl to your Coupa instance so supplier data, approval chains, and users sync automatically.
📍In Compyl, go to Integrations › Coupa and enter your instance URL, client ID, client secret, and (optionally) a form filter, then authorize via OAuth and save.
— Names exact config fields and includes setup steps.Nov 12th, 2026
0200
Authentication required on workflow-engine APIsbreaking60how completely this was documenteddepth20/40specificity25/30actionability15/3060 / 100
Internal workflow-engine API endpoints now require authentication and return 401 to unauthenticated callers.
— Names the response code but not specific endpoint paths.Nov 12th, 2026
thinner coverage below
0300
Expanded sync coverage for Netskope and AWS integrationsimproved55how completely this was documenteddepth20/40specificity25/30actionability10/3055 / 100
The Netskope integration now collects DLP incidents, SCIM users, groups, and group memberships. The AWS integration now syncs GuardDuty detector status, threat intelligence sets, and IP sets.
— Names synced data types but no setup path given.Nov 12th, 2026
0400
Searchable custom fields on Controls and Clients gridsimproved55how completely this was documenteddepth20/40specificity20/30actionability15/3055 / 100
Custom fields on the Controls and Clients grids can now be searched and filtered, matching the behavior already available on Risks, Vendors, and other record types.
— Names affected grids but no exact filter syntax.Nov 12th, 2026
0500
Multiple third-party links on incidentsnew30how completely this was documenteddepth10/40specificity10/30actionability10/3030 / 100
You can now link an incident to multiple third parties directly from the incident record.
— Brief statement with no detail on mechanism or UI.Nov 12th, 2026
Runlayer is an MCP gateway that enforces identity-mapped permissions and detects threats across the agents and MCP servers employees use.
Runlayer's 5.1.0 release adds an aiwatch enforce-test verification flow, OpenCode session monitoring with Protect/Enforce coverage, policy-aware agent connection UI, and broader redaction and error-masking across AI Watch — alongside a breaking change requiring RESEND_API_KEY for self-hosted managed email.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
└──▷ WHAT SHIPPED ·
6 FEATURES0 OF 6 FEATURESmost completely described firstwhat's the number?
Each feature carries 0–100 for how completely the vendor documented
it — not how big or important the work is. A major capability
described in eight words scores low, and that is the finding.
depth0–40what it does and how it works or what changed
specificity0–30names real surfaces — APIs, flags, formats, limits, numbers
actionability0–30enough to go use it — a named endpoint, flag, or config key tops this; a UI path is a starting point
Code and config rank first by construction: a documented endpoint,
flag or config key scores at the top of specificity and actionability,
so it sorts above a feature described only in prose. Under
60 the notes go thin — everything below the “thinner
coverage” line is thinner documentation, not smaller work.
Hover any meter for that feature's three sub-scores.
0100
RESEND_API_KEY required for managed email deliverybreaking95how completely this was documenteddepth35/40specificity30/30actionability30/3095 / 100
Managed email delivery (Let Runlayer handle email delivery) for self-hosted deployments now reads exclusively from RESEND_API_KEY, with no fallback to WorkOS Vault for its key.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄ Set it in the provisioned tenant secret on ECSECSElastic Common Schema, an open specification maintained by Elastic that defines a standard field naming convention for log and event data, letting cyber tools share and correlate records without per-source field mapping. or in the runlayer-app Secret on Kubernetes before upgrading, or organizations with managed delivery will fall back to custom SMTPSMTPThe application-layer protocol used to send email between servers and from clients to servers; cyber tools use it to relay alerts, phishing simulations, or exfiltrated data over standard mail infrastructure. or send no email at all, including security-violation alerts.
— Exact env var and secret locations with upgrade consequence spelled out.5.1.0 (2026-11-13)
0200
Enforce test detections via aiwatch enforce-testnew85how completely this was documenteddepth30/40specificity25/30actionability30/3085 / 100
Run aiwatch enforce-test on a device to print a test MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface. server entry and two shell prompts that Enforce and Protect block, then confirm each block in the agent reply and in Shadow AI Activity. A remote MCP entry pointed at the Runlayer-hosted test server lets teams verify Enforce blocks remote MCP calls on devices that have no enforce-test command yet.
After a policy rollout, verify that Enforce and Protect are blocking on a device by printing the test prompts and checking results in Shadow AI Activity.
$ aiwatch enforce-test
— Runnable CLI command with exact verification steps and location.5.1.0 (2026-11-13)
0300
OpenCode session monitoring in AI Watchnew70how completely this was documenteddepth30/40specificity25/30actionability15/3070 / 100
AI Watch now records OpenCode sessions and applies Protect and Enforce to OpenCode tool calls on macOS and Windows, including MCP tools, via a Runlayer-installed OpenCode plugin (OpenCode 1.x, 1.18.29 or newer recommended; OpenCode 2 not yet supported).
— Names versions and platforms but no install command given.5.1.0 (2026-11-13)
0400
Policy-aware agent connection UInew60how completely this was documenteddepth25/40specificity20/30actionability15/3060 / 100
Agent account connection cards now show deny policies that reach the agent, noting how many tools a policy denies and marking the connector Blocked by policy for blanket denies; denied tools appear first in the tool picker and cannot be selected.
Error masking for local MCP server errorsimproved60how completely this was documenteddepth25/40specificity20/30actionability15/3060 / 100
Error messages returned by local MCP servers run through the Runlayer CLI are now masked when PIIPIIPersonally Identifiable Information — data that can identify a specific individual, such as names, email addresses, or SSNs. Cyber tools flag or redact it to limit exposure and meet privacy regulations. detection or token masking is set to mask, preventing personal data and credentials in errors from reaching the AI client.
— Names the `mask` setting but not its exact config key.5.1.0 (2026-11-13)
thinner coverage below
0600
Redaction in AI Watch events and audit exportsimproved50how completely this was documenteddepth20/40specificity20/30actionability10/3050 / 100
AI Watch skill discovery and risk audit events, and audit log exports, now show ***REDACTED*** in place of sign-in details, query values, and recognized tokens in a skill's git remote URL.
— Names the redaction marker but no control to configure it.5.1.0 (2026-11-13)
└──▷ BREAKING ON UPGRADE
!
Managed email delivery ('Let Runlayer handle email delivery') for self-hosted deployments no longer falls back to WorkOS Vault for its key — `RESEND_API_KEY` must be set in the ECS tenant secret or the `runlayer-app` Kubernetes Secret before upgrading, or organizations with managed delivery will fall back to custom SMTP or send no email at all, including security-violation alerts.Our best guess
doberman-core adds AUTH-prompt integration for Claude Code, Codex and Cursor in doberman dash, a sandbox 2FA check in doberman doctor, single-use TOTPTOTPA one-time password algorithm (RFC 6238) that generates short-lived codes from a shared secret and the current time, giving cyber tools phishing-resistant second-factor authentication without a server round-trip. enforcement, a dedicated judge API key variable, and a module-invocation fallback.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
└──▷ WHAT SHIPPED ·
5 FEATURES0 OF 5 FEATURESmost completely described firstwhat's the number?
Each feature carries 0–100 for how completely the vendor documented
it — not how big or important the work is. A major capability
described in eight words scores low, and that is the finding.
depth0–40what it does and how it works or what changed
specificity0–30names real surfaces — APIs, flags, formats, limits, numbers
actionability0–30enough to go use it — a named endpoint, flag, or config key tops this; a UI path is a starting point
Code and config rank first by construction: a documented endpoint,
flag or config key scores at the top of specificity and actionability,
so it sorts above a feature described only in prose. Under
60 the notes go thin — everything below the “thinner
coverage” line is thinner documentation, not smaller work.
Hover any meter for that feature's three sub-scores.
0100
Judge API key via `DOBERMAN_JUDGE_API_KEY`new75how completely this was documenteddepth25/40specificity25/30actionability25/3075 / 100
The optional judge now prefers the DOBERMAN_JUDGE_API_KEY environment variable, falling back to ANTHROPIC_API_KEY if unset, letting the judge use a credential scoped separately from the shared Anthropic key.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
Point the optional judge at a dedicated API key instead of the shared Anthropic key so credentials stay scoped.
— Names both env vars and gives a runnable examplev0.20.0
0200
Module invocation via `python -m doberman`new65how completely this was documenteddepth22/40specificity18/30actionability25/3065 / 100
doberman-core can now be run without the CLI binary on PATH using python -m doberman, for example python -m doberman doctor, useful in restricted environments such as CI containers where the binary isn't installed.
Invoke doberman when the doberman binary is missing from PATH, such as inside a restricted CI container.
$ python -m doberman doctor
— Runnable command given, but limited context beyond thatv0.20.0
thinner coverage below
0300
`doberman doctor` sandbox 2FA checknew45how completely this was documenteddepth15/40specificity15/30actionability15/3045 / 100
doberman doctor now checks whether Claude Code's sandbox keeps your 2FA secret isolated from the agent.
— Names the check but not its mechanism or outputv0.20.0
0400
Single-use TOTP codesimproved45how completely this was documenteddepth25/40specificity10/30actionability10/3045 / 100
TOTP codes are now single-use: a reused or older code is refused even while it is still inside its validity window.
— Explains behaviour change but no config surface to act onv0.20.0
0500
`doberman dash` AUTH-prompt integration for Claude Code, Codex, Cursornew35how completely this was documenteddepth10/40specificity15/30actionability10/3035 / 100
doberman dash adds AUTH-prompt integration for Claude Code, Codex and Cursor.
— Only named in the summary, no mechanism or usage detailv0.20.0
ToolHive adds workload-identity-based client authentication via SPIFFESPIFFEAn open standard, maintained by the CNCF, for assigning cryptographic identities to workloads in dynamic infrastructure so services can authenticate each other without passwords or static credentials.X.509-SVIDsX.509-SVIDsX.509 certificates formatted to carry a SPIFFE Verifiable Identity Document, encoding a workload's cryptographic identity as a URI SAN so services can mutually authenticate without passwords or API keys., cross-authorization-server token delegation through ID-JAG token minting, and trace correlation for audit events.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
└──▷ WHAT SHIPPED ·
3 FEATURES2 OF 3 FEATURESmost completely described firstwhat's the number?
Each feature carries 0–100 for how completely the vendor documented
it — not how big or important the work is. A major capability
described in eight words scores low, and that is the finding.
depth0–40what it does and how it works or what changed
specificity0–30names real surfaces — APIs, flags, formats, limits, numbers
actionability0–30enough to go use it — a named endpoint, flag, or config key tops this; a UI path is a starting point
Code and config rank first by construction: a documented endpoint,
flag or config key scores at the top of specificity and actionability,
so it sorts above a feature described only in prose. Under
60 the notes go thin — everything below the “thinner
coverage” line is thinner documentation, not smaller work.
Hover any meter for that feature's three sub-scores.
0101
SPIFFE X.509-SVID client authenticationnew45how completely this was documenteddepth20/40specificity15/30actionability10/3045 / 100
ToolHive can now authenticate clients using SPIFFE X.509X.509A standard defining the format of public-key certificates, maintained by the ITU-T, used by cyber tools to authenticate identities and establish encrypted connections over TLS and similar protocols.-SVIDsSVIDsShort-lived, cryptographically verifiable identity documents issued by the SPIFFE framework to workloads, giving cyber tools a standard way to authenticate services to each other without managing certificates manually., enabling workload-identity-based access without relying on bearer tokensbearer tokensA credential that grants access to whoever presents it, without a separate proof of possession. Protecting the token matters because a stolen copy can be used by someone else..︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
— Describes the mechanism but no config keys or flags givenv0.65.0
0202
ID-JAG token minting and actor-token validationnew45how completely this was documenteddepth20/40specificity15/30actionability10/3045 / 100
ToolHive now mints ID-JAG tokens and validates actor tokens for delegate clients, supporting cross-authorization-server token delegation workflows.
— Explains workflow but lacks concrete API or flag namesv0.65.0
0300
Trace and span IDs on audit eventsimproved30how completely this was documenteddepth15/40specificity10/30actionability5/3030 / 100
ToolHive now records trace and span IDs on audit events, correlating audit log entries with distributed traces.
— States the change but no schema or field names givenv0.65.0
Minimal is a secure, declarative runtimes for Claude Code, Gemini, Codex, Pi and more, so you can run multiple agents safely on your stack from a verified software supply chain.
Minimal added a tamper-evident audit trail for boxes with parent-chain verification surfaced in min doctor, plus several smaller CLI additions for inspecting and managing boxes.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
└──▷ WHAT SHIPPED ·
4 FEATURES0 OF 4 FEATURESmost completely described firstwhat's the number?
Each feature carries 0–100 for how completely the vendor documented
it — not how big or important the work is. A major capability
described in eight words scores low, and that is the finding.
depth0–40what it does and how it works or what changed
specificity0–30names real surfaces — APIs, flags, formats, limits, numbers
actionability0–30enough to go use it — a named endpoint, flag, or config key tops this; a UI path is a starting point
Code and config rank first by construction: a documented endpoint,
flag or config key scores at the top of specificity and actionability,
so it sorts above a feature described only in prose. Under
60 the notes go thin — everything below the “thinner
coverage” line is thinner documentation, not smaller work.
Hover any meter for that feature's three sub-scores.
0100
JSON/YAML output for box detailsnew65how completely this was documenteddepth20/40specificity20/30actionability25/3065 / 100
Box details, including its execs, can now be output as JSON or YAML with min box show -o json or min box show -o yaml.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
— Exact flag and formats given; no deeper mechanism explainedrelease-dd7d8cf2
0200
Strip terminal control codes from box logsimproved60how completely this was documenteddepth20/40specificity20/30actionability20/3060 / 100
min box logs now strips terminal control codes when stdout is not a terminal.
— Clear before/after condition and command namedrelease-dd7d8cf2
thinner coverage below
0300
List and show box typesnew40how completely this was documenteddepth10/40specificity15/30actionability15/3040 / 100
New min box list and show commands let users list and show box types.
The code-oss project is an open-source code editor combining editing, debugging, and extensibility for developers across Windows, macOS, and Linux.
CVE-2026-102230: Information Disclosure via Integrated Browser File Uploads (CVSSCVSSCommon Vulnerability Scoring System, a method for describing a vulnerability's characteristics and severity. Its score is not the probability that attackers will exploit it. 7.4, affects < 1.145.3)︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
CVE-2026-121662: Workspace Trust Bypass via Prototype Pollution in Workspace Configuration (CVSSCVSSCommon Vulnerability Scoring System, a method for describing a vulnerability's characteristics and severity. Its score is not the probability that attackers will exploit it. 8.6, affects < 1.145.3)︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
CVE-2026-121660: Security Feature Bypass via Inherited Publisher Names in Allowed Extensions Policy (CVSSCVSSCommon Vulnerability Scoring System, a method for describing a vulnerability's characteristics and severity. Its score is not the probability that attackers will exploit it. 7.5, affects < 1.145.3)︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
CVE-2026-120782: Command Execution via Agent-Originated Local Commands (CVSSCVSSCommon Vulnerability Scoring System, a method for describing a vulnerability's characteristics and severity. Its score is not the probability that attackers will exploit it. 6.8, affects < 1.145.3)︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
CVE-2026-120780: Remote Code Execution via serve-web Release Downgrade (CVSSCVSSCommon Vulnerability Scoring System, a method for describing a vulnerability's characteristics and severity. Its score is not the probability that attackers will exploit it. 7.2, affects < 1.145.3)︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
Cloudflare is a connectivity cloud platform delivering internet security, performance, and developer infrastructure services.
Cloudflare's November update adds spending limits and alert thresholds for Workers and Durable Objects, expands D1 and AI Search integration (including a breaking namespace change in the Terraform provider), extends RealtimeKit with token validity windows and noise cancellation, and open-sources its Kitesurf browser engine for AI agents.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
└──▷ WHAT SHIPPED ·
5 FEATURES0 OF 5 FEATURESmost completely described firstwhat's the number?
Each feature carries 0–100 for how completely the vendor documented
it — not how big or important the work is. A major capability
described in eight words scores low, and that is the finding.
depth0–40what it does and how it works or what changed
specificity0–30names real surfaces — APIs, flags, formats, limits, numbers
actionability0–30enough to go use it — a named endpoint, flag, or config key tops this; a UI path is a starting point
Code and config rank first by construction: a documented endpoint,
flag or config key scores at the top of specificity and actionability,
so it sorts above a feature described only in prose. Under
60 the notes go thin — everything below the “thinner
coverage” line is thinner documentation, not smaller work.
Hover any meter for that feature's three sub-scores.
0100
Custom RealtimeKit token validity windows and revocationnew90how completely this was documenteddepth30/40specificity30/30actionability30/3090 / 100
Set custom validity windows for RealtimeKit auth tokens by passing not_before and expires_on in the token object when adding a participant via POST /client/v4/accounts/$ACCOUNT_ID/realtime/kit/$APP_ID/meetings/$MEETING_ID/participants; individual tokens can now be revoked without deactivating the whole meeting.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
Issue a RealtimeKit participant token that is valid only during a scheduled session window, so a shared link stops working after the session ends.
AI Search namespace support and breaking Terraform changesbreaking90how completely this was documenteddepth35/40specificity30/30actionability25/3090 / 100
Terraform provider v5.29.0 adds a namespace argument to the cloudflare_ai_search_instance resource, enabling management of AI Search instances in any namespace, and introduces cloudflare_ai_search_namespace as a new resource. Breaking: cloudflare_ai_search_instance import IDs now require the namespace segment (<account_id>/<namespace>/<id>, use default for the default namespace), the cloudflare_ai_search_instance and cloudflare_ai_search_instances data sources now require namespace, the filter.namespace argument on the cloudflare_ai_search_instance data source is removed, cloudflare_ai_search_instances now returns instances from one namespace only, and changing namespace on cloudflare_ai_search_instance replaces the instance.
Spending limits for Workers and Durable Objectsnew80how completely this was documenteddepth30/40specificity30/30actionability20/3080 / 100
Set spending limits and email alert thresholds for Workers and Durable Objects via the Cloudflare dashboard under Spend limits, or with the cf CLI using billing spend-policies with --product-id, --threshold-cents, and --action usage_block.
Noise cancellation for RealtimeKit web appsnew70how completely this was documenteddepth25/40specificity30/30actionability15/3070 / 100
Add noise cancellation to RealtimeKit web apps with the new @cloudflare/realtimekit-noise-cancellation package, calling RealtimeKitNoiseCancellation.init({ meeting }) after meeting initialisation. Works in React, Angular, and plain JavaScript on Chrome, Edge, Firefox, and Safari, including mobile.
Local testing for Browser Run quickActionimproved70how completely this was documenteddepth25/40specificity25/30actionability20/3070 / 100
Browser Run .quickAction() calls can now be tested locally under npx wrangler dev, where the browser binding runs against a local Chrome instance instead of requiring remote mode.
The `cloudflare_ai_search_instance` import ID now requires the namespace segment: `<account_id>/<namespace>/<id>`. For instances in the default namespace, use `<account_id>/default/<id>`.Our best guess
The `cloudflare_ai_search_instance` and `cloudflare_ai_search_instances` data sources now require a `namespace` argument; set `namespace = "default"` for instances in the default namespace.Our best guess
The `filter.namespace` argument on `cloudflare_ai_search_instance` data source is removed; `cloudflare_ai_search_instances` now returns instances from one namespace only.Our best guess