THE CYBER TOOLCHAIN
NO. 125
The daily firehose — everything the toolchain shipped today, already filtered.
We read every release from the 1309 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.
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.
This issue came out before editions were ranked, so it shows the Standard edition in issue order.
$ tct list # 57 tools matched
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. Aembit's kill switch cuts one user's 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. client session dead without knocking over every other agent on the tenant. GitLab's Orbit graph goes GA, so you can ask which services a vulnerability actually reaches instead of grepping. Snyk now runs its header, cookie and TLS checks against every Extra Host, not just the primary URL.
Aembit revokes a named user's MCP client session on demand, targeted or in bulk and callable from the API, so a suspected agent compromise doesn't force a tenant-wide credential rotation. Runlayer 2.0.0 closes cross-session manipulation and IP-spoofing gaps in the same layer, which is what makes per-session revocation meaningful rather than theatre.
Orbit is now GA for Premium and Ultimate: a live graph linking code, pipelines, vulnerabilities and ownership, queryable in plain language or by exact regex, with MCP wired to the CLI. Blast-radius questions that previously meant tracing imports by hand and chasing down a CODEOWNERS file become a single query.
Snyk API & Web applies security header, cookie attribute and 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./TLS checks to every Extra Host on a target — the vhosts and alternate domains that quietly diverge from the primary config. Prowler's AWS security group check now catches rules that only become open exposure when their CIDRsCIDRsClassless Inter-Domain Routing notation: a compact way to express an IP address range as a base address plus a prefix length (e.g., 192.168.1.0/24), letting cyber tools specify network scopes concisely. are combined, a class of misconfiguration that per-rule evaluation reads as clean.
Sysdig's preview automations collapse Vulnerability Findings notifications into one per resource or image rather than one per finding, which is the difference between a ticket queue and a channel nobody reads. Checkmarx's Fusion SASTSASTStatic application security testing, analyzing source code or compiled code for vulnerabilities without running the application. It helps developers find problems before deployment. ships with Triage, Remediation and Developer Assist to push the same compression into the fix step, generating patches and surfacing them in the IDE.
Fast service fingerprinting CLI for 170+ protocols (TCP/UDP/SCTP) - built by Praetorian
$ go install github.com/praetorian-inc/nerva/cmd/nerva@v1.84.0
$ git clone --branch v1.84.0 https://github.com/praetorian-inc/nerva.git # already have the repo? check out this version: $ git checkout v1.84.0
Go module path matches the repo · checked 30 Oct
Get started with Semgrep to help you catch, flag, and fix real vulnerabilities before they ship.
Semgrep's latest release adds a new related-issues API endpoint and direct Anthropic AI provider connectivity, expands Supply Chain dependency coverage and filtering, adds PowerShell support to Guardian, and ships several breaking changes including removal of QL language support and a renamed xAI provider.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
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.
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.
A new experimental endpoint, GET /api/deployments/{deployment_id}/issues/{issue_id}/related-issues, lets you query a finding's related issues directly.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄ The existing relatedIssues field on finding responses is now deprecated in favor of this endpoint.
$ curl -H 'Authorization: Bearer <token>' 'https://semgrep.dev/api/deployments/<deployment_id>/issues/<issue_id>/related-issues'
Support for QL (the CodeQL query language) is removed: --lang ql and the ql language key in rules are now rejected, and .ql and .qll files are no longer parsed.
Semgrep Supply Chain now includes every package declared in a Python requirements.txt file — including packages with version ranges (e.g. requests>=2.0) and packages without a version (e.g. flask) — in the dependency inventory and SBOMSBOMA software bill of materials, an inventory of the components and versions inside an application. It helps teams find affected software when a dependency has a vulnerability or licensing issue., not only packages pinned with ==.
AI provider settings now include a required Base model field for Azure OpenAI deployments, needed for new Azure OpenAI setups. Organizations that had not chosen an AI provider mode now default to the Cloudflare Workers AI mode, while organizations that chose their own providers or are on the minimal data retention policy remain unchanged.
The xAI AI provider is renamed to SpaceXAI and its endpoint field is removed; the provider always uses https://api.x.ai/v1.
Rate limits for /api/v2/deployments/public/{id} and /api/v2/deployments/public/search now count every request including repeated and concurrent requests.
The Advisories page gains Language, Severity, and Advisory type filters, matching the filters already available on the Findings page, plus a new Incident column in the advisories table.
Dependencies can now be filtered by version directly on the Dependencies page without opening advanced options; the dependency filter is now an All dependencies menu with Exact and Contains matching modes and All versions or Select versions options.
Selected Agentic Workflows issues can now be exported to CSV; Semgrep exports only the issues or groups selected, or uses active filters when nothing is selected, with a 10,000-issue limit applied to the selection.
A sortable False positive rate column is added to the rules table in Policies, covering Semgrep Code and Semgrep Secrets rules, with the ability to select a rate to open associated false-positive findings.
Users can connect directly to Anthropic by supplying their own Anthropic API key under the custom AI provider mode in 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 settings, for use with in-app features. Available to Semgrep Enterprise and existing bring-your-own-key users.
A Finding type filter is added to dashboards to view AI-powered and Rule-based findings separately in charts, drilldowns, and PDF exports.
Semgrep Guardian now supports PowerShell, running its hooks on both PowerShell commands and Bash commands.
--lang ql and the ql language key in rules are now rejected, and .ql and .qll files are no longer parsed.xAI AI provider is renamed to SpaceXAI and its endpoint field is removed; the provider always uses https://api.x.ai/v1./api/v2/deployments/public/{id} and /api/v2/deployments/public/search now count every request including repeated and concurrent requests.Veracode is a cloud-based application security platform that scans code, binaries, and running applications to find and help remediate software vulnerabilities.
Veracode integrations add 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. policy and predependency command parameters to veracode.yml across GitHub, GitLab, and Azure DevOps workflows, environment setup checks in VS Code, and platform-managed GitHub install and scan scheduling.
predependency_command parameter to veracode.yml in the GitHub, GitLab, and Azure DevOps Workflow Integrations to run commands before Veracode packages the project.build_runs_on parameter to veracode.yml in the GitHub Workflow Integration for SCASCASoftware composition analysis, identifying third-party dependencies and checking them for known vulnerabilities and license issues. It covers software a team uses rather than writes itself. scans.policy parameter to veracode.yml in the GitHub, GitLab, and Azure DevOps Workflow Integrations for IaC scans.veracode.yml in the Azure DevOps Workflow Integration.break_build_policy_findings and break_build_on_error are both set to true.11 details in Everything
-repourl is explicitly specified.GitHub expanded its Meta API with new webhook IP ranges that will break static allowlists if not updated, and brought Copilot plugin management to general availability with repository-level defaults and enterprise-wide controls.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
GitHub is a web-based platform for version control and collaborative software development using Git repositories.
GitHub expanded its Meta API with new webhook IP ranges that will break static allowlists if not updated, and brought Copilot plugin management to general availability with repository-level defaults and enterprise-wide controls.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
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.
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.
The Meta API's hooks field now includes three new CIDRCIDRA notation for a block of IP addresses, such as 10.0.0.0/24. The number after the slash says how many leading address bits identify the network. prefixes — 9.234.98.160/28, 9.234.98.176/28, and 9.234.106.48/28 — covering all github.com plans including GitHub Enterprise Cloud.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄ Static network allowlists for incoming webhook deliveries must be updated to include these ranges or they will reject the new traffic.
$ curl -s https://api.github.com/meta | jq '.hooks'
Repository maintainers can now define default plugins and a plugin marketplace so contributors get consistent Copilot plugin setup without repeating configuration per project, and contributors can turn off those repository-default plugins where enterprise policy permits. Enterprise administrators can require plugins (auto-installed and user-locked), block plugins, and restrict installation to approved marketplaces through managed settings that override repository and personal choices. The controls apply across VS Code, Copilot CLI, and the GitHub Copilot app, with managed plugin controls available on GitHub Enterprise Cloud for Copilot Business and Copilot Enterprise licenses, and repository defaults also available to individual Copilot users.
9.234.98.160/28, 9.234.98.176/28, and 9.234.106.48/28 unless those CIDR prefixes are added manually.JetBrains Qodana’s official command line tool
qodana-cli adds automatic exchange of an organization API token for a project-scoped token, narrowing the credential scope needed for CI scans.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
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.
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.
qodana-cli now automatically exchanges an organization API token for a project-scoped token, reducing the credential scope needed for CI scans.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
Checkmarx introduces Fusion SASTSASTStatic application security testing, analyzing source code or compiled code for vulnerabilities without running the application. It helps developers find problems before deployment., an AI-augmented scanning engine, alongside a trio of AI Assist capabilities — Triage Assist, Remediation Assist, and Developer Assist — that automate finding triage, fix generation, and in-IDE remediation.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
Checkmarx agentic application security software platform Combines Hybrid scanning, AI-powered agents, and unified risk intelligence across every attack surface
Checkmarx introduces Fusion SASTSASTStatic application security testing, analyzing source code or compiled code for vulnerabilities without running the application. It helps developers find problems before deployment., an AI-augmented scanning engine, alongside a trio of AI Assist capabilities — Triage Assist, Remediation Assist, and Developer Assist — that automate finding triage, fix generation, and in-IDE remediation.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
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.
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.
Remediation Assist automatically generates AI remediation suggestions after a scan or automated triage completes, with admins defining the projects, branches, scanners, and severity thresholds in scope.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄ Project overrides let individual teams adapt these automated remediation settings where permitted.
Fusion SAST combines established rule-based security analysis with frontier-model reasoning, and a Findings Analysis Engine consolidates results to reduce false positives and duplicate findings. Checkmarx reports an F1 score of 0.74 versus 0.64 for the Next-Gen SAST foundation.
Triage Assist (AI Triage) evaluates whether a finding is reachable and exploitable, automatically setting a finding's state to Confirmed or Proposed Not Exploitable, with its reasoning visible in Risk Orchestration.
Developer Assist brings security analysis and remediation guidance into IDE and AI coding workflows, including supported autonomous workflows where it works with the coding agent to repeat remediation and rescanning.
Snyk Open Source is a developer security tool that scans project dependencies for known vulnerabilities and license issues, suggesting fixes or upgrades to remediate them.
Snyk API & Web now extends its security header, cookie attribute, and 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./TLS checks to every Extra Host configured on a target, not just the primary target URL.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
The official changelog for the Snyk AI Trust Platform. Discover what's new in our security intelligence, AI-powered tools, and latest product enhancements.
Snyk API & Web now extends its security header, cookie attribute, and 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./TLS checks to every Extra Host configured on a target, not just the primary target URL.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
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.
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.
Snyk API & Web now applies security header, cookie attribute, and SSL/TLS checks — including HSTS header not enforced and Missing Content Security Policy header — to every Extra Host configured on a target, not just the primary target URL endpoint.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄ As a result, scans of targets with Extra Hosts configured may report new findings for those hosts on their next scan even if nothing on the target itself has changed.
Teams and agents ship trusted software faster with GitLab, the intelligent orchestration platform with context and governance for the entire software lifecycle.
GitLab's Orbit context graph reaches general availability for Premium and Ultimate customers, giving agents and engineers a live, queryable graph of code, pipelines, vulnerabilities and ownership — paired with new ways to query it, from CLI-wired 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. agents to plain-language vulnerability blast-radius hunts and exact regex code search.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
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.
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.
MCP-compatible agents (Claude Code, Codex, Cursor, OpenCode, Gemini CLI, Antigravity) connect to Orbit by running glab auth login then glab orbit setup; custom agents and internal tooling connect through REST; agents on GitLab's Duo Agent Platform query Orbit natively with nothing to configure; and engineers can query the graph directly through the Data Explorer.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
$ glab auth login
glab orbit setup
GitLab Orbit, a live queryable graph connecting code, merge requests, pipelines, vulnerabilities, work items, and ownership across all projects in a group, is now generally available for Premium and Ultimate customers on GitLab.com, with Self-Managed and GitLab Dedicated support coming in the following weeks and a free 30-day trial available through account teams. Orbit indexes every push, merge, and deploy in real time — at GitLab scale, more than 1.5 million projects and 8 TB of graph data in under 2 hours — so agents reason from current lifecycle state. Orbit access follows existing GitLab roles so people and agents see only what they are permitted to see, and the graph queries behind each answer can be audited. Orbit is sold as query packs (a shared pool per org) priced at $30 in GitLab Credits per user per month for 1,000 queries, with overages billed at 33 queries per GitLab Credit (~$0.03 per query).
Security teams can ask GitLab Duo Agentic Chat plain-language questions against the Orbit graph to find blast radius without manual correlation, for example: 'Use Orbit to map the blast radius of PHP in this group. List every project that contains PHP code, grouped by subgroup, with owners.' GitLab's built-in Security Analyst Agent now uses Orbit directly to traverse the graph and answer these questions, and teams can also collect every container scanning finding for specific CVEs across a group via Agentic Chat (requires Ultimate and Security Manager role).
At the top-level group, exact code search with regular expression mode lets engineers find things like container base images alongside Orbit graph queries, e.g. lang:Dockerfile ^FROM.*php or file:composer\.json$.
$ lang:Dockerfile ^FROM.*php
Detect secrets + live validation. Map blast radius.
Kingfisher's v2.16.0 release expands its built-in rule catalog to 501 rules with new connection-string password detectors, adds detection of bare AWS access IDs, and updates its Betterleaks integration with new content detectors.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
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.
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.
Five new connection-string password rules now ship by default, covering PostgreSQL/MySQL/MariaDB JDBCJDBCA Java standard API for connecting applications to relational databases, maintained as part of the Java SE platform, letting tools query any compliant database through a single consistent interface. query passwords and SQL Server JDBC properties.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄ They are embedded by default in the CLI, kingfisher-rules, and the Python SDK, bringing the total rule catalog to 501 rules.
Kingfisher now detects AWS access IDs even when no secret key is found nearby, though validation and revocation still require a matching secret key.
The Betterleaks integration now reflects v2.0.0-rc.5, adding eight new content detectors for Labelbox, NuGetNuGetMicrosoft's package manager for .NET, hosting a public registry of reusable libraries so tools can pull in dependencies without bundling or manually maintaining third-party code., and Tailscale.
GitGuardian SaaS is a cloud-hosted security platform that detects exposed secrets, API keys, and credentials in source code repositories and developer workflows.
GitGuardian SaaS adds automatic GitHub team sync into workspaces, extends custom webhooks to support multiple headers, ships new Public API endpoints for incident/ticket management, tightens Endpoint Protection scope grants in ggshield, and broadens Bridge routing for host-derived secrets.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
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.
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.
Enabling team synchronization on a GitHub integration makes GitGuardian mirror the GitHub organization's teams, members, and repository ownership into the workspace.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄ Sync runs automatically every 24 hours or on demand via Sync now, with results shown in a synchronization history view.
The ggshield auth login authorization page now groups Endpoint Protection scopes into a single opt-in card and no longer grants them by default; existing automation that relied on those scopes being pre-selected will need to re-authenticate.
GitGuardian Bridge now routes validity checks and secret analysis for detectors where the host is read from the secret itself (such as database connection strings) whenever that host is mapped to a bridge — previously this routing was limited to detectors with a custom host configured.
Custom webhook integrations now accept multiple custom headers per webhook — such as 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. or routing headers — configured from the webhook's settings page.
New Public API endpoints let you unlink a ticket from an incident, and retrieve the stored content of an occurrence of an internal or public incident in clear text — the latter subject to the same access requirements as secret access.
ggshield auth login — users must explicitly opt in via the new single card, so existing automation that relied on those scopes being pre-selected will need to re-authenticate.GitGuardian Detection Engine scans source code repositories, commits, and pipelines to detect leaked secrets such as API keys, passwords, and credentials.
Zarf v0.89.0 introduces the v1beta1 package format with major schema changes — granular SBOMSBOMA software bill of materials, an inventory of the components and versions inside an application. It helps teams find affected software when a dependency has a vulnerability or licensing issue./attachment layers, an object-based differential field, and removed ###ZARF builtins — alongside custom OCIOCIOpen Container Initiative standard, maintained by the Linux Foundation, defining the image format and runtime spec that container tools share so images built once run anywhere OCI-compliant. credential providers, granular layer selection in the zoci Load API, and several breaking changes including opt-in kustomizekustomizeA Kubernetes-native configuration management tool, maintained by the Kubernetes project, that layers and patches YAML manifests without templating, letting cyber tools ship reusable base configs that operators override cleanly. plugins, a removed cachepath SDK option, and removed OCI reference shortcut aliases.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
The Airgap Native Package Manager for Kubernetes
Zarf v0.89.0 introduces the v1beta1 package format with major schema changes — granular SBOMSBOMA software bill of materials, an inventory of the components and versions inside an application. It helps teams find affected software when a dependency has a vulnerability or licensing issue./attachment layers, an object-based differential field, and removed ###ZARF builtins — alongside custom OCIOCIOpen Container Initiative standard, maintained by the Linux Foundation, defining the image format and runtime spec that container tools share so images built once run anywhere OCI-compliant. credential providers, granular layer selection in the zoci Load API, and several breaking changes including opt-in kustomizekustomizeA Kubernetes-native configuration management tool, maintained by the Kubernetes project, that layers and patches YAML manifests without templating, letting cyber tools ship reusable base configs that operators override cleanly. plugins, a removed cachepath SDK option, and removed OCI reference shortcut aliases.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
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.
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.
v1beta1 introduces a new init package format for cluster initialization, splits SBOM and documentation into separate granular files and layers (the documentation field is renamed to attachments), changes differential from a scalar to an object, improves component config override import logic for more predictable composition, removes ###ZARF builtin substitutions in favor of explicit values, adds an id field to the image object in the package schema, and validates annotation size and package secret size at zarf package create time to reject oversized values early; package data is no longer duplicated in the package secret.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
zarf now supports granular layer selection from the zoci Load API, letting callers fetch only the layers they need instead of requiring all layers; existing code relying on the previous all-layers default will break.
The global cachepath SDK option is removed; cache paths must now be set per-operation, and image layer caching is simpler and no longer shared across concurrent operations.
Kustomize plugins are no longer active by default on read-only dev commands; users must now explicitly opt in via a flag rather than having them active by default.
zarf now supports custom credential providers for OCI registries, enabling non-standard auth flows when pushing or pulling packages.
Defense-unicorn shortcut aliases (short-form OCI refs) are removed from OCI package references; full references are now required.
Proxy mode now requires an injector image to be specified explicitly; deployments that omitted it will fail.
Release candidate versions now pass zarf version requirement checks.
zoci Load API now requires callers to specify granular layers; existing code that relied on the previous all-layers default will break.zarf package create).v1beta1, differential is now an object; configs using it as a scalar will be rejected.v1beta1, the documentation field is renamed to attachments; any package definition using documentation must be updated.v1beta1, annotations exceeding the new size limit will be rejected at create time.id field is added to the image object; tools parsing the image schema must handle the new field.###ZARF builtin substitutions are removed from v1beta1; packages relying on them must be rewritten to use explicit values.v1beta1, SBOM and documentation are now stored as separate granular files and layers; tooling that reads the old combined format will break.cachepath SDK option is removed; any SDK consumer setting it must switch to per-operation cache configuration.Endor Labs shipped an Antigravity Plugin Marketplace integration that brings its 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 AURI Agents into Google Antigravity's agentic development workflows, adding real-time dependency and secret scanningsecret scanningChecking files, code or other data for exposed credentials such as passwords, API keys and tokens. Findings help teams remove the secret and revoke or rotate it. plus automated remediation, decision-support, and vulnerability-explanation agents.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
Endor Labs is an application security platform built for AI-generated and human-written code, helping teams prioritize and fix real risk across the SDLC.
Endor Labs shipped an Antigravity Plugin Marketplace integration that brings its 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 AURI Agents into Google Antigravity's agentic development workflows, adding real-time dependency and secret scanningsecret scanningChecking files, code or other data for exposed credentials such as passwords, API keys and tokens. Findings help teams remove the secret and revoke or rotate it. plus automated remediation, decision-support, and vulnerability-explanation agents.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
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.
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.
The SCASCASoftware composition analysis, identifying third-party dependencies and checking them for known vulnerabilities and license issues. It covers software a team uses rather than writes itself. Remediation AURI Agent runs inside Antigravity on the developer's own infrastructure using their own LLMLLMA large language model, a neural network trained on large amounts of text to recognize patterns and generate language. Applications use it for tasks such as answering, summarizing and writing code. keys and source-control credentials.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄ Triggered with a prompt such as 'Use the SCA Remediation agent to fix the vulnerable dependency in this project,' it applies reachability and upgrade impact analysis, selects the safest upgrade, opens a pull request, and attaches evidence to the PR body.
The Vulnerability Explainer AURI Agent turns a CVE ID into a plain-language explanation of impact and remediation steps specific to the project. Asking 'Use the Vulnerability Explainer agent to explain CVE-2024-12345 in the context of this project' returns a project-specific impact summary and recommended remediation steps.
Endor Labs added a one-click install from Google Antigravity's Plugin marketplace that brings the Endor Labs MCP server and AURI Agents (including AI SASTSASTStatic application security testing, analyzing source code or compiled code for vulnerabilities without running the application. It helps developers find problems before deployment. Triage) into Antigravity-based agentic workflows. The MCP server gives Antigravity agents real-time security context — checking packages for known vulnerabilities and malicious code before a dependency is added, scanning for secrets, and surfacing vulnerability details inline while code is written.
The Dependency Decision Helper AURI Agent answers whether a package should be added before an agent installs it, running inside Antigravity on the developer's own infrastructure.
StepSecurity's Harden Runner adds a flag to block implicit AWS endpoint allowances on self-hosted Linux runners, exposes new eBPFeBPFA Linux technology for running verified programs inside the kernel without changing its source. Tools use it to observe processes, enforce rules and process network traffic. monitoring health metrics for ARC-hosted runners, and extends support to Avrea-hosted enterprise runners.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
StepSecurity is a security platform that hardens CI/CD pipelines, such as GitHub Actions, by detecting supply-chain risks, enforcing least-privilege runners, and monitoring workflow tampering.
StepSecurity's Harden Runner adds a flag to block implicit AWS endpoint allowances on self-hosted Linux runners, exposes new eBPFeBPFA Linux technology for running verified programs inside the kernel without changing its source. Tools use it to observe processes, enforce rules and process network traffic. monitoring health metrics for ARC-hosted runners, and extends support to Avrea-hosted enterprise runners.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
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.
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.
ARC Harden Runner v2.24.0 reports eBPF monitoring health metrics — ring buffer usage, dropped network/DNS/process events, and DNS tracking map inserts, removes, and errors — via a Prometheus endpoint when env.enableMetrics is enabled.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄ Buffer sizing is configurable via keys such as dnsTrackingMapSize and processRingBufferSize.
env: dnsTrackingMapSize: 65536 processRingBufferSize: 32768 enableMetrics: true
A new disable_implicit_endpoints feature flag stops self-hosted Linux runners from implicitly allowing AWS endpoints, while GitHub and StepSecurity endpoints remain allowed. The flag shipped in both Self-Hosted VM Harden Runner v1.9.5 and GitHub Action v2.22.3 (enterprise tier).
env: disable_implicit_endpoints: true
Harden Runner now supports runners hosted by Avrea (enterprise tier), added in GitHub Action v2.22.3.
Invicti's latest release adds client certificate secrets for target authentication, enriches the API catalog with sensitive-data labels and scanning filters plus discovery notifications, and introduces a Scan CLI environment variable to halt DASTDASTDynamic application security testing, probing a running application from the outside for vulnerabilities. It checks behavior exposed through the application's interfaces. scans on policy failure.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
Official documentation for Invicti AppSec Platform, Invicti Platform, and Invicti Enterprise — web application security testing, DAST, API security, and vulnerability management.
Invicti's latest release adds client certificate secrets for target authentication, enriches the API catalog with sensitive-data labels and scanning filters plus discovery notifications, and introduces a Scan CLI environment variable to halt DASTDASTDynamic application security testing, probing a running application from the outside for vulnerabilities. It checks behavior exposed through the application's interfaces. scans on policy failure.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
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.
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.
The API catalog and API discovery now label each API and endpoint with the category of sensitive data it exposes — Secrets, 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., or Financials — so risk is visible without opening each finding.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄ A new Requires scanning filter surfaces newly discovered or recently updated APIs that still need a DAST scan. Users also receive bell notifications and emails when APIs are discovered or updated, with a direct link to the changed APIs in the catalog; email alerts can be turned off in profile notification settings.
A new INVICTI_FAIL_FAST environment variable stops DAST scans on first policy failure when run via the Scan CLI.
Client certificate files can now be stored as secrets and selected in a target's client certificate authentication settings, rather than being configured per-target directly.
Cortex Analyzers Repository
Cortex-Analyzers 3.10.2 enhances the ANY.RUN sandbox analyzer with proxy and domain flexibility, and adds artifact support to the analyzer testing framework.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
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.
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.
The ANY.RUN analyzer now supports proxy configuration, alternative domains, and opt_automated_interactivity for sandbox submissions.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
TestAnalyzer now supports artifacts, allowing artifact-based observable testing during analyzer development.
Xage Security is a company providing zero-trust access control and data protection for industrial operational technology and critical infrastructure, enabling secure remote access, identity-based policy enforcement, and encrypted data sharing.
Xage adds a SailPoint Identity Security integration for bidirectional sync of User and User Group entitlements, automated provisioning, and revocation.
3 details in Everything
Malcolm is a powerful, easily deployable network traffic analysis tool suite for full packet capture artifacts (PCAP files), Zeek logs and Suricata alerts.
Malcolm v26.10.0 adds normalized 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. threat-intelligence fields across every intel source, Community IDCommunity IDA vendor-neutral network flow hashing standard, maintained by Corelight, that produces a consistent identifier for a connection across different tools so analysts can correlate events from disparate data sources. computation in LogstashLogstashAn open-source data pipeline daemon maintained by Elastic that ingests, transforms, and routes log and event data, giving cyber tools a flexible way to normalize and forward high-volume telemetry without custom plumbing., expanded OT protocol port and field mapping, and a richer document-lookup API.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
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.
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.
New ZEEK_PORTS_… environment variables register non-default ports for thirteen ICSNPP OT protocol analyzers (BACnetBACnetA communication protocol for building automation systems, standardized as ASHRAE 135, covering HVAC, lighting, and access control devices. Cyber tools use it to discover and interact with building infrastructure on IP or serial networks., C12.22, EtherNet/IP, GE-SRTP, Genisys, HART-IP, Modbus, Omron FINS, OPC UA Binary, Profinet IO-CM, ROC+, S7comm, and Synchrophasor), so a site running these protocols on nonstandard ports no longer needs a custom Zeek script.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄ As part of this, the older ZEEK_SYNCHROPHASOR_PORTS, ZEEK_GENISYS_PORTS, and ZEEK_ENIP_PORTS variables are renamed to ZEEK_PORTS_SYNCHROPHASOR, ZEEK_PORTS_GENISYS, and ZEEK_PORTS_ENIP; existing values are migrated on upgrade, but external automation referencing the old names must be updated.
ZEEK_PORTS_MODBUS=5020/tcp
Threat intelligence enrichment now maps into normalized ECS threat.* fields across every source: Zeek intel kill chain phases map to threat.tactic.*; MISP MITREMITREA US federally funded research and development center that maintains public frameworks—ATT&CK, CVE, D3FEND—that cyber tools use to classify threats, vulnerabilities, and defensive techniques in a common language. galaxy tags in cif_tags map to threat.technique.* and threat.technique.subtechnique.*; Mandiant and Google Threat Intelligence actors map to threat.group.name; and Suricata Emerging Threats mitre_* metadata and capa subtechnique results are mapped the same way, with ATT&CKATT&CKA publicly maintained knowledge base by MITRE cataloguing adversary tactics, techniques, and procedures observed in real attacks, giving cyber tools a common taxonomy for classifying and correlating threat behavior. for ICSICSIndustrial Control Systems — hardware and software that monitor and control physical processes like power grids, pipelines, and manufacturing lines, making them a target for cyber tools focused on operational technology environments. distinguished from Enterprise. Per-indicator metadata is also carried into Zeek intel, including Mandiant categories, campaigns, reports, opt-in threat rating (meta.threat_score), and STIX 2.1STIX 2.1A JSON-based open standard maintained by OASIS for representing threat intelligence objects and relationships, letting cyber tools exchange indicators, actors, and TTPs in a common machine-readable format. confidence values, with Mandiant's M-Score treated as confidence and absent values left unset.
Logstash can now optionally compute Community ID for Zeek logs that arrive without one, controlled by the LOGSTASH_ZEEK_COMPUTE_COMMUNITY_ID environment variable, enabled by default. An ID Zeek already supplied is never overwritten, and transport is inferred by log type for single-transport analyzers.
LOGSTASH_ZEEK_COMPUTE_COMMUNITY_ID=true
The Malcolm API's document lookup gains sort, offset, and track_total_hits arguments and a total hit count with its exact-or-lower-bound relation; both the document lookup and field aggregations endpoints now return shards and timed_out status so a client can distinguish a partial or timed-out search from a genuinely empty result. Argument validation errors return HTTP 400 with an error string, and setting MALCOLM_API_DEBUG=true logs a compact per-request search summary.
Zeek OT protocol fields are normalized to Arkime field names for BACnet, C12.22, DNP3DNP3A serial and TCP/IP communication protocol used in industrial control systems, especially power grids and water utilities, to let control software poll and command field devices like sensors and switches., EtherNet/IP (ENIP/CIP), Modbus, OPC UA, S7comm, and Synchrophasor so filters and drill-downs carry across Dashboards and Arkime; original zeek.* fields are preserved alongside the new top-level fields.
Zeek is updated to 9.0.1 LTS, parsing and indexing two new log types: multicast_participants.log, surfaced in a new Multicast dashboard, and ntp_control.log and ntp_private.log, added to the index template and Arkime field definitions.
Hedgehog Linux ISOISOA widely adopted international standards body that publishes technical specifications; cyber tools cite ISO standards to signal compliance with recognized benchmarks for quality, security, or interoperability. installation now prompts whether to run the sensor status kiosk at startup, defaulting to yes; choosing no disables kiosk.service and removes the browser autostart entry.
New ioc_types and include_autogenerated_tags parameters for Google Threat Intelligence collections let a deployment load only the indicator types it needs and omit machine-generated tags, reducing memory cost in Zeek.
NETBOX_DEVICETYPE_IMPORT_VENDORS limits the NetBoxNetBoxAn open-source network infrastructure management platform maintained by the NetBox Labs community, used to model IP addresses, devices, and topology so cyber tools can query a live source of truth about the environment. device-type library import to selected manufacturers on first import, also constraining which manufacturers are matched during MAC OUI enrichment.
Two Zeek detections authored by Corelight are added: php2shell-detect for the php2shell vulnerabilities and emudhra-weak-keys for weak eMudhra certificate keys.
ZEEK_SYNCHROPHASOR_PORTS, ZEEK_GENISYS_PORTS, and ZEEK_ENIP_PORTS are renamed to ZEEK_PORTS_SYNCHROPHASOR, ZEEK_PORTS_GENISYS, and ZEEK_PORTS_ENIP; existing values are migrated on upgrade, but any external automation referencing the old names must be updated.Asimily is the proactive cyber asset defense platform for IoT, OT, IT, and IoMT devices, with prioritized risk mitigation.
Asimily Edge now deploys as a Cisco IOx container on Catalyst switches for deep packet inspection of east-west traffic without a separate appliance.
4 details in Everything
Sophos Central Firewall adds asynchronous processing for SD-WANSD-WANSoftware-defined wide-area networking: an approach that decouples WAN routing logic from physical hardware, letting operators centrally control traffic paths across MPLS, broadband, and LTE links to reduce cost and improve resilience. connection group operations, along with a Force Delete option for stuck groups.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
Sophos Central Firewall is a cloud-managed network security solution that provides firewall protection, intrusion prevention, and threat monitoring for organizations, administered through the Sophos Central platform.
Sophos Central Firewall adds asynchronous processing for SD-WANSD-WANSoftware-defined wide-area networking: an approach that decouples WAN routing logic from physical hardware, letting operators centrally control traffic paths across MPLS, broadband, and LTE links to reduce cost and improve resilience. connection group operations, along with a Force Delete option for stuck groups.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
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.
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.
Create, edit, and delete operations on SD-WAN connection groups are now processed asynchronously in the background, with progress visible on the SD-WAN Connection and Task Queue pages.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄ A new Force Delete option lets admins remove connection groups that cannot be deleted through the normal delete flow.
Plerion is a cloud security platform that detects misconfigurations, vulnerabilities, and threats across cloud environments and provides remediation guidance.
Plerion adds SASTSASTStatic application security testing, analyzing source code or compiled code for vulnerabilities without running the application. It helps developers find problems before deployment. scanning for GitLab and Azure DevOps integrations and a new AWS SHIP lens on the Well-Architected page.
2 details in Everything
Prowler 5.47.0 adds new Azure compliance checks for Application Gateway 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./HTTP2 and subscription owner counts, annotates MCP ServerMCP ServerA process that exposes tools, resources, and prompts over Anthropic's Model Context Protocol, letting an AI agent call external APIs or local services through a standardised interface. tools with read/write intent hints, redesigns the Issues list around exploitability, unifies the add-provider wizard across all supported clouds, and improves AWS security group exposure detection to catch combined-CIDRCIDRA notation for a block of IP addresses, such as 10.0.0.0/24. The number after the slash says how many leading address bits identify the network. rules.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
This repo hosts Prowler OSS (CLI and Local Server). Prowler is the world’s most widely used open-source cloud security platform that automates security and compliance across any cloud environment for Humans and AI Agents.
Prowler 5.47.0 adds new Azure compliance checks for Application Gateway 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./HTTP2 and subscription owner counts, annotates MCP ServerMCP ServerA process that exposes tools, resources, and prompts over Anthropic's Model Context Protocol, letting an AI agent call external APIs or local services through a standardised interface. tools with read/write intent hints, redesigns the Issues list around exploitability, unifies the add-provider wizard across all supported clouds, and improves AWS security group exposure detection to catch combined-CIDRCIDRA notation for a block of IP addresses, such as 10.0.0.0/24. The number after the slash says how many leading address bits identify the network. rules.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
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.
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.
Prowler adds network_application_gateway_waf_enabled and network_application_gateway_http2_enabled checks for Azure Application Gateways and their WAF policies, plus iam_subscription_owners_between_2_and_3 which verifies each subscription has between two and three principals with the Owner role, including owners inherited from management groups.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄ All three new checks are mapped to CISCISA nonprofit that publishes consensus-based configuration benchmarks and controls for hardening operating systems, cloud environments, and software; cyber tools use it as a ready-made compliance baseline. 5.0 and 6.0.
Prowler's AWS security group internet exposure checks now treat CIDRs in a single ingress rule that together cover 0.0.0.0/0 or ::/0 (such as 0.0.0.0/1 plus 128.0.0.0/1) as open to the internet.
Prowler's 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 now annotates every tool with readOnlyHint and destructiveHint so MCP clients can distinguish read-only listing calls from writes; mute rule and mutelist writes are marked destructive.
Prowler's add-provider wizard now connects every provider — AWS, Azure, GCP, Microsoft 365, GitHub, Alibaba Cloud, and Cloudflare — in one step, with teammate invitations available for all providers and Lighthouse AI assistance beside the wizard.
Prowler's Issues list now leads with the verdict, shows the attack pathattack pathA possible sequence of steps an attacker could take to reach a valuable target. It connects weaknesses, permissions and reachable systems rather than treating each finding separately. under each title, and adds a Why column with five evidence legs; the menu entry is renamed to Exploitability issues.
Sysdig Secure adds preview automations that group Vulnerability Findings notifications by resource or image instead of sending one per finding.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
Sysdig Secure is a cloud security platform that provides runtime threat detection, vulnerability management, and compliance monitoring for containers, Kubernetes, and cloud workloads.
Sysdig Secure adds preview automations that group Vulnerability Findings notifications by resource or image instead of sending one per finding.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
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.
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.
Sysdig Secure now supports Vulnerability Findings automations that group findings by resource or image, sending a single Slack or email notification — or creating a single Jira ticket — per resource or image rather than per finding.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄ This feature is in Preview.
CLI to install, manage & troubleshoot Kubernetes clusters running Cilium
cilium-cli v0.20.2 improves connectivity testing through restrictive proxies and namespace labeling, while removing deprecated BGPBGPA routing protocol that lets autonomous systems on the internet exchange reachability information, determining how traffic flows between networks. Cyber tools use it to monitor route hijacks or inject sinkholes. CRDCRDA Kubernetes extension that adds a new object type to the cluster, so a tool's configuration is stored, validated and versioned by Kubernetes itself rather than in a file beside it. support.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
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.
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.
The deprecated v2alpha1 version of the BGP CRDsCRDsA Kubernetes extension that adds a new object type to the cluster, so a tool's configuration is stored, validated and versioned by Kubernetes itself rather than in a file beside it. is removed; clusters using v2alpha1 BGP CRDs must migrate before upgrading.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
cilium-cli now falls back to 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. when a proxy rejects the SPDYSPDYA deprecated Google-developed network protocol designed to speed up HTTP by multiplexing streams, compressing headers, and prioritizing requests over a single TCP connection; it became the basis for HTTP/2. upgrade, allowing connectivity tests to work through proxies that block SPDY.
cilium-cli now applies labels to all new test namespaces, making it easier to identify and filter test resources.
v2alpha1 version of the BGP CRDs is removed; clusters using v2alpha1 BGP CRDs must migrate before upgrading.security-profiles-operator v1.1.2 adds a spoc sign subcommand with SLSA Build L3SLSA Build L3A supply-chain integrity framework level defined by the OpenSSF SLSA specification, requiring builds to run in a hardened, isolated environment so the build platform itself can guarantee no human or process tampered with the artifact. provenanceprovenanceEvidence recording where a software artifact came from and how it was built, including its source and build process. It helps a consumer check that the artifact matches a trusted build. for its bundle, catalog, and security profiles, plus a new SPOD option to reject Permissive-mode SELinuxSELinuxA mandatory access control system built into the Linux kernel, maintained by the NSA and Red Hat, that enforces fine-grained policies restricting what processes can access, even if they run as root. profiles, alongside several breaking changes to metrics auth, Helm validation, profile policy, and resource naming.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
The Kubernetes Security Profiles Operator
security-profiles-operator v1.1.2 adds a spoc sign subcommand with SLSA Build L3SLSA Build L3A supply-chain integrity framework level defined by the OpenSSF SLSA specification, requiring builds to run in a hardened, isolated environment so the build platform itself can guarantee no human or process tampered with the artifact. provenanceprovenanceEvidence recording where a software artifact came from and how it was built, including its source and build process. It helps a consumer check that the artifact matches a trusted build. for its bundle, catalog, and security profiles, plus a new SPOD option to reject Permissive-mode SELinuxSELinuxA mandatory access control system built into the Linux kernel, maintained by the NSA and Red Hat, that enforces fine-grained policies restricting what processes can access, even if they run as root. profiles, alongside several breaking changes to metrics auth, Helm validation, profile policy, and resource naming.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
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.
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.
SelinuxProfiles that run in Permissive mode — which bypass the SELinux denylist — can now be rejected cluster-wide by setting spec.selinux.options.allowPermissiveProfiles: false in the SPOD.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
spec:
selinux:
options:
allowPermissiveProfiles: false
Metrics access moves from the default ServiceAccount to the new spo-metrics-client ServiceAccount; scrapers using the metrics-token Secret must switch to spo-metrics-client-token or obtain a token via kubectl create token spo-metrics-client.
security-profiles-operator ships SLSASLSASupply-chain Levels for Software Artifacts, a framework for assessing how software was built and protecting that process from tampering. Provenance records help verify where a build came from. Build L3 provenance for the bundle, catalog, and security profiles, and adds a new spoc sign subcommand to sign OCIOCIOpen Container Initiative standard, maintained by the Linux Foundation, defining the image format and runtime spec that container tools share so images built once run anywhere OCI-compliant. artifacts (e.g. spoc sign <oci-artifact-ref>) after recording them, for supply-chain verification.
$ spoc sign <oci-artifact-ref>
SelinuxProfiles that use the CIL keywords all, and, not, or, range, or xor as permissions are now rejected.
New SecurityProfilesOperatorDaemons must be named spod; other names are now rejected.
The Helm chart now rejects unknown values, so previously-ignored typos or stale keys in Helm chart configuration will now cause install/upgrade failures.
default ServiceAccount to the new spo-metrics-client ServiceAccount; scrapers using the metrics-token Secret must switch to spo-metrics-client-token or use kubectl create token spo-metrics-client.all, and, not, or, range, or xor as permissions are rejected.spod.Sophos Anti-Virus for macOS is a security product that scans Mac systems for malware, viruses, and other threats, providing real-time protection and remediation.
Sophos Anti-Virus for macOS gained AI Defense integration via Sophos Fusion and a new Web Profiles system with Gen AI categorization that replaces classic web control.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
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.
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.
Sophos Anti-Virus for macOS now supports Web Profiles with Gen AI categorization, replacing the previous classic web control system.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
Sophos Anti-Virus for macOS now supports AI Defense through Sophos Fusion, providing visibility and control over AI usage.
SentinelOne's Singularity platform protects endpoints, cloud, identity and data with AI-driven detection, Purple AI and an agentic SOC layer.
SentinelOne adds Singularity ASOCASOCApplication Security Orchestration and Correlation — a category of platforms that aggregate findings from multiple testing tools, deduplicate and prioritize them, and automate remediation workflows so cyber tools can be managed at scale., a new autonomous SOCSOCA security operations center, the team or function that monitors threats, investigates alerts and coordinates incident response. Its work can be internal, outsourced or shared. suite with agentic investigation, hyperautomation, AI 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., IOCIOCAn indicator of compromise, an observable clue that may point to malicious activity, such as a suspicious file hash or domain. It is evidence to investigate, rather than proof by itself. management, and a credits-based consumption model.
2 details in Everything
Trend Vision One is Trend Micro's unified cybersecurity platform that consolidates threat detection, response, and risk management across endpoints, cloud, email, and networks.
Trend Vision One's Agentless Vulnerability & Threat Detection update lets AWS customers keep agentless scanning traffic inside their own VPCVPCA logically isolated virtual network inside a public cloud provider's infrastructure, letting tools deploy resources with private addressing, controlled routing, and firewall rules without managing physical hardware. instead of routing through public endpoints.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
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.
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.
Agentless Vulnerability & Threat Detection can now be deployed so scanning traffic stays within your own VPC rather than routing through public endpoints.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄ In the console, go to Cloud Security → Cloud Accounts → AWS and redeploy using the CloudFormationCloudFormationAn AWS service that lets you describe and provision cloud infrastructure as declarative YAML or JSON templates, so a cyber tool can reproducibly spin up or tear down its entire AWS environment with a single API call. template, selecting your target VPC and subnet when prompted.
Databricks shipped GA dashboard relationships for multi-fact AI/BI data models, a round of Genie Agents UX and performance improvements, Beta support for running multiple real-time flows per serverless pipeline alongside adaptive source-stage repartitioning for Kafka/Kinesis, and new ug usage CLI commands for spend-limit visibility.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
Databricks is a cloud-based data lakehouse platform for data engineering, analytics, machine learning, and AI workloads, built on Apache Spark.
Databricks shipped GA dashboard relationships for multi-fact AI/BI data models, a round of Genie Agents UX and performance improvements, Beta support for running multiple real-time flows per serverless pipeline alongside adaptive source-stage repartitioning for Kafka/Kinesis, and new ug usage CLI commands for spend-limit visibility.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
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.
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.
View up to five personal and shared spend limits sorted by highest percentage used with ug usage, or see all applicable limits with ug usage --all.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
$ ug usage --all
Source stage adaptive repartitioning automatically adjusts the number of tasks reading from Kafka or Kinesis sources to match traffic, removing the need to manually set maxPartitions or shardsPerTask.
Genie Agents gains a guided creation experience that walks through building an effective agent, retrieves files faster when answering questions (Beta), hides content search details such as the number of indexed files from users with CAN VIEW or CAN RUN permission on a Genie Agent (Beta), and now opens links to volume files in an embedded agent directly instead of redirecting to the workspace.
You can now run multiple real-time mode flows within a single serverless pipeline without the flows competing for task slots; this is available on the preview channel in Beta.
Dashboard relationships are now generally available, letting you build complex multi-fact data models scoped to an AI/BI dashboard.
Sumo Logic is a cloud-native machine data analytics platform that collects, searches, and analyzes logs and metrics for security monitoring, troubleshooting, and operational insights.
Sumo Logic introduced a self-service Manage Features capability, letting paid customers toggle organization-level features on or off directly from the console.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
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.
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.
Paid customers can now opt in or out of specific Sumo Logic features for their organization directly from the Organizations UI, without needing to raise a support request.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄ In the console, go to Organizations UI > Manage Features, then toggle the desired feature on or off.
Cribl Stream is a vendor-neutral observability pipeline that routes, filters, transforms, and reduces machine data from logs and metrics before sending it to analytics or storage destinations.
Cribl Stream adds a Microsoft 365 Copilot Source and a Microsoft Fabric Real-Time Intelligence destination for telemetry collection and routing.
2 details in Everything
Open-Source Unified Vulnerability Management, DevSecOps & ASPM
DefectDojo's generic parser now supports Package URLPackage URLA standardized string format for uniquely identifying software packages across ecosystems (npm, PyPI, Maven, etc.), letting cyber tools reference and correlate dependency data without ecosystem-specific logic. (purl) identifiers for dependency findings, enabling direct links between findings and specific packages.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
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.
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.
The generic parser now accepts a Package URL (purl) for a finding's dependency, letting you link findings directly to a specific package.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
OXO is a security scanning orchestrator for the modern age.
oxo adds support for scanning Autonomous System Numbers as a new asset type, alongside a minor protobufprotobufA language-neutral binary serialization format created by Google that encodes structured data more compactly and faster than JSON or XML, making it attractive for high-throughput inter-process or network communication. enhancement for TestFlight assets.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
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.
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.
oxo can now scan Autonomous System Numbers directly using the new oxo scan run asn subcommand and the asn asset type, allowing recon scans to enumerate infrastructure associated with an ASNASNAutonomous System Number — a globally unique identifier assigned by regional internet registries to a network operator, letting routing protocols distinguish which organization controls a block of IP addresses..︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
$ oxo scan run --install --agent agent/ostorlab/nmap asn AS15169
oxo's TestFlight asset now includes an app name field in its protobuf definition.
Automox is moving exclusion window management from device groups to individual devices, adding a device-level UI workflow while deprecating the group-scoped scheduled window API endpoints.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
Automox is a check out our full list of release notes, and stay up to date with what
Automox is moving exclusion window management from device groups to individual devices, adding a device-level UI workflow while deprecating the group-scoped scheduled window API endpoints.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
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.
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.
As exclusion management moves to the device level, Automox is updating or deprecating these public API endpoints: Create a New Scheduled Window, Get a Scheduled Window, Update a Scheduled Window, Delete a Scheduled Window, List Scheduled Windows (search), 'Check Whether a Scheduled Window is Currently Active', Check Scheduled Window Status for Server Groups, List Next Scheduled Window Dates by Group, and List Next Scheduled Window Dates by Device.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄ New per-device exclusion window API endpoints will replace these group-scoped endpoints for programmatic management.
Admins can now assign and manage exclusion windows directly on individual devices via a new Edit Devices action in the Automox console, replacing the group-association workflow. Device tables and Device Details pages now show which exclusion windows apply to each specific device. As a behavior change, moving a device between groups no longer adds or removes its exclusion window, so any group-based scripts or onboarding flows that rely on group placement to apply or remove exclusion windows must be updated to manage exclusion windows as a separate, independent step.
Iru is a device-centric platform that unifies security, IT, and compliance around devices and everything they touch.
Iru Agent for Windows 1.25.7 strengthens the integrity of its download pipeline and improves app inventory visibility on Windows devices.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
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.
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.
Iru Agent for Windows now validates the checksum of every app and update it downloads, installing only files that match what Iru sent.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
App inventory for Windows computers now reports the install path for more applications.
Microsoft Intune is a cloud-based unified endpoint management solution built on Zero Trust. Empower IT to manage, secure and govern devices, apps and endpoints.
Microsoft Intune adds bulk eSIMeSIMA programmable SIM embedded directly in a device's hardware, standardized by the GSMA, allowing carrier profiles to be provisioned or switched remotely without swapping a physical card. activation for up to 100 corporate-owned Android EnterpriseAndroid EnterpriseA Google program that gives organizations APIs and management modes to enroll, configure, and enforce policy on Android devices through an EMM platform. devices, with eSIM data plan preservation during bulk wipes.
Devices > All devices > Bulk device actions, supplying a carrier activation server URL, on Android 15 or later (COBO, COSU, and COPE form factors).2 details in Everything
An asset discovery and inventory platform that identifies and maps all devices on networks without requiring agents.
runZero adds has_owner search to the vulnerability inventory and avoids probing Allen-Bradley/Rockwell 1785-ENET devices on identification.
has_owner search term.2 details in Everything
Axonius is a cybersecurity asset management platform that aggregates device and user data from many security and IT tools to find coverage gaps and enforce security policies.
Axonius 10.0.1 adds AI Workforce auto-pause, a new Adapter Noise Filters Worker, AVS factor querying, Security Finding Risk Score v2, and scoped Vulnerability Repository access.
7 details in Everything
LoginRadius: Privacy-compliant identity and access management platform. Offer Authentication and Authorization for your web and mobile apps in just a few clicks, with: SSO, MFA, social logins and more.
LoginRadius enhanced its Admin Console with direct Phone ID management for user records and tightened permission restrictions for view-only users.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
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.
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.
Admins can add, update, or remove a user's Phone ID directly from User details > Edit User Info.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄ The field only appears when phone login is enabled, and changing the Phone ID marks it as unverified.
View-only users are now restricted from creating or editing items under Organizations and Custom IdP > Settings.
JumpCloud's biggest window moves were a new 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. integration for Directory Insights, expanded Apple MDMMDMMobile device management, centrally configuring and managing enrolled devices. Administrators use it to apply settings, distribute applications and enforce device security policies. Device Lock with automatic PIN escrow, a Scrut Automation integration for continuous audit evidence, and a broad Agentic 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. release adding non-human identitynon-human identityA non-human identity, an account or credential used by software or a device rather than a person. Service accounts, workload identities and automation tokens need their own permissions and lifecycle controls. lifecycle management, continuous audit trails, real-time risk detection, Conditional-Access device trust, workflows automation, and an upcoming PAMPAMBrokering administrator sessions through a gateway instead of handing out standing credentials. Access is granted for one session, to one target, and recorded. module.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
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 biggest window moves were a new 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. integration for Directory Insights, expanded Apple MDMMDMMobile device management, centrally configuring and managing enrolled devices. Administrators use it to apply settings, distribute applications and enforce device security policies. Device Lock with automatic PIN escrow, a Scrut Automation integration for continuous audit evidence, and a broad Agentic 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. release adding non-human identitynon-human identityA non-human identity, an account or credential used by software or a device rather than a person. Service accounts, workload identities and automation tokens need their own permissions and lifecycle controls. lifecycle management, continuous audit trails, real-time risk detection, Conditional-Access device trust, workflows automation, and an upcoming PAMPAMBrokering administrator sessions through a gateway instead of handing out standing credentials. Access is granted for one session, to one target, and recorded. module.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
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.
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.
The Device Lock MDM command for macOS now automatically generates and escrows the lock PIN, retrievable from Admin Portal under Highlights > Device Lock > View PIN.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄ Admins can show a custom message and phone number on the Lock Screen for Mac, iPhone, and iPad, and track lock status from the Devices list via the new macOS Lock Status column.
JumpCloud now integrates with Scrut Automation so that active identity and device settings across Windows, Mac, and Linux endpoints are automatically collected and mapped to compliance framework requirements, removing manual screenshot-gathering during audits. AI agent ownership and access tracking also feeds into Scrut to produce auditor-ready evidence that agents follow required safety rules, across 70+ compliance frameworks.
Continuous audit trails now link every human and non-human action to a verified identity, supporting SOC 2SOC 2An auditing framework from the American Institute of CPAs that certifies a service organization's controls around security, availability, and data handling, giving buyers assurance that a vendor meets baseline trust criteria., ISOISOA widely adopted international standards body that publishes technical specifications; cyber tools cite ISO standards to signal compliance with recognized benchmarks for quality, security, or interoperability. 27001, HIPAAHIPAAA US federal law enacted in 1996 that sets privacy and security requirements for protected health information, giving cyber tools a compliance baseline to enforce or audit against., and NIS2NIS2EU directive that expands mandatory cybersecurity requirements to more sectors and member states than its predecessor, giving cyber tools a compliance framework to map controls and reporting obligations against. audits.
Device trust now blocks unmanaged or unhealthy devices from accessing company data using Conditional AccessConditional AccessA policy enforcement framework, most commonly Microsoft's in Entra ID, that grants or blocks resource access based on signals like user identity, device state, and location — letting cyber tools trigger step-up auth or deny risky sessions. Policies (CAP).
Agentic IAM provides full visibility and lifecycle tracking across AI agents and non-human identitiesnon-human identitiesA non-human identity, an account or credential used by software or a device rather than a person. Service accounts, workload identities and automation tokens need their own permissions and lifecycle controls., automatically updating agent access when a user changes roles or leaves.
Privileged access managementPrivileged access managementBrokering administrator sessions through a gateway instead of handing out standing credentials. Access is granted for one session, to one target, and recorded. (PAM) will deploy isolated identities and time-boxed credentials for admins and agents, coming December 2026.
Risk detection flags risky sign-ins, non-human behavior, and unapproved access changes in real time.
Workflows automation now automates lifecycle processes, offboarding, and policy remediation across all identities and devices.
All the latest releases from the 1Password team
1Password CLI 2.43.0-beta.02 adds a command to list secrets stored in Environments, alongside a small documentation fix for file attachment reads.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
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.
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.
The new op environment secret list command, added in CLI 2.43.0-beta.02, reads secrets from Environments, letting users audit stored values or pipe them into scripts.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
$ op environment secret list
The op read command's help text now includes an example demonstrating how to read a file attachment.
ConductorOne's single release this window adds role mining suggestions that proactively identify candidate groups for role creation and let analysts automate membership enrollment for new hires.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
ConductorOne is an identity governance platform that automates workforce access management and deploys AI agents to streamline access reviews and provisioning.
ConductorOne's single release this window adds role mining suggestions that proactively identify candidate groups for role creation and let analysts automate membership enrollment for new hires.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
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.
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.
ConductorOne surfaces role mining suggestions showing each group's size, shared entitlements, and a confidence score based on attributes such as department, job title, or manager.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄ Selecting a suggestion opens an analysis page with the group's filters pre-filled, letting reviewers filter by user attribute or by access held and limit scope to specific apps; reviewers can then turn on membership automation so users matching the group's filters — including new hires — receive the role's access automatically. Any earlier role mining analysis can be resumed from the start page.
Secure users, AI agents, and more with Auth0, an easy-to-implement, scalable, and adaptable authentication and authorization platform.
Auth0 brings RFC 9421RFC 9421An IETF standard defining HTTP Message Signatures, a scheme for cryptographically signing HTTP requests and responses so tools can verify message integrity and authenticity end-to-end. HTTP Message Signatures to Tenant Network ACLsACLsAn access control list, a set of rules allowing or denying operations for particular users, services or network addresses. It decides who can use a resource., giving enterprise tenants cryptographic verification of request origin at the network boundary.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
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.
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.
Auth0 Tenant Network ACLs now support RFC 9421 HTTP Message Signatures via the Management API, letting enterprise tenants enforce HMAC-SHA256HMAC-SHA256A keyed hash that proves a message came from someone holding a shared secret and was not altered. Webhooks and API request signing commonly use it. request verification at the network boundary to block direct tenant bypass.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄ Incoming requests can be validated to originate only from authorized reverse proxies or gateways, with unverified direct traffic dropped. Signing keys can be rotated with zero downtime by configuring multiple active signing keys per rule, and key-rotation activity and verification events can be observed via acls_summary Tenant Log events.
Clerk is an authentication and user-management platform that provides pre-built UI components, APIs, and SDKs for adding sign-up, sign-in, and identity features to applications.
Clerk now supports passkeyspasskeysPhishing-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. as a second factor, allowing users to complete multi-factor authenticationmulti-factor authenticationMulti-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. with a 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. after an initial sign-in method.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
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.
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.
Clerk now supports passkeys as a second factor for multi-factor authentication, letting users verify with a passkey after signing in with another method.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
Okta identity engine is a secure, scalable, and highly available authentication and user management for any app.
Okta Identity Engine adds 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 OAuth client and Cross App Access support, High Risk Proxy detection, 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. AAGUIDAAGUIDA 128-bit identifier embedded in a FIDO2/WebAuthn authenticator that identifies the make and model of the device, letting relying parties enforce authenticator policy without tracking individual users.-only group matching, and new Privileged Access System Log events.
9 details in Everything
DigiCert CertCentral added API-level visibility into supported intermediate CA chains and automated key-type handling for certificate orders through its Europe Services API.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
DigiCert CertCentral is a web-based platform for ordering, issuing, managing, and renewing SSL/TLS certificates and other digital trust credentials.
DigiCert CertCentral added API-level visibility into supported intermediate CA chains and automated key-type handling for certificate orders through its Europe Services API.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
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.
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.
When ordering through the CertCentral Europe Services API (https://certcentral.digicert.eu/services/v2), if the optional ca_cert_id parameter is omitted on a product set to the DigiCert default ICA chain, the system automatically performs key-type matching between the ICA chain and the submitted CSRCSRA Certificate Signing Request is a block of encoded data submitted to a certificate authority, containing a public key and identity details, so the CA can issue a signed TLS certificate without ever seeing the private key..︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
The Product info API now exposes an allowed_ca_certs field listing the intermediate CA (ICA) chains supported for a given product. If the field is absent from the response, ICA selection is not enabled for that product.
Tailscale introduced a new route selection policy to replace Regional Routing, added per-machine posture assertion visibility in the admin console, and shipped a Terraform breaking change alongside a minor macOS UI fix.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
Tailscale creates private WireGuard-based mesh networks with centralized identity and access controls.
Tailscale introduced a new route selection policy to replace Regional Routing, added per-machine posture assertion visibility in the admin console, and shipped a Terraform breaking change alongside a minor macOS UI fix.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
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.
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.
The split_dns field in the tailscale_dns_configuration Terraform resource is now defined as a set instead of a list, which may require updates to existing configurations.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
A new Route selection setting in the admin console replaces the Regional Routing toggle, letting admins control how the tailnet chooses between redundant subnet routers, app connectors, exit nodes, and Tailscale Service hosts. Configure it via Settings > Route selection to set MagicRoute preferences.
The postures page of the admin console now shows posture assertion details per machine, including each machine's value, pass/fail status, and a count of policy file rules that require a given posture.
MacBooks with a notched display now show a notification window when the Tailscale menu bar icon is hidden behind the notch.
split_dns field in the tailscale_dns_configuration Terraform resource is now defined as a set, not a list, which may require updates to existing configurations.A scalable overlay networking tool with a focus on performance, simplicity and security
Nebula v1.12.0 adds Linux UDPUDPA network protocol that sends individual messages without guaranteeing delivery or order. Applications trade built-in reliability for less overhead and can add their own recovery./TUNTUNA virtual network device in Linux (and other OSes) that lets a userspace program read and write raw IP packets, making it the standard building block for VPNs and traffic-manipulation tools. offloads, CPUCPUThe main general-purpose processor in a computer. It runs application instructions and coordinates work with memory and other hardware. thread pinning, a local control API reachable via nebula ctl, an HTTP info socket for node identity, FIPS 140-3FIPS 140-3A US federal standard, maintained by NIST, that defines requirements for cryptographic modules used in government systems. Cyber tools target it to satisfy procurement mandates requiring validated crypto. build mode, and experimental split DNS, alongside a large set of breaking changes to config validation, SSH debug server behavior, and the embedder API.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
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.
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.
Nebula can offload packet processing to the Linux kernel via tun.use_offloads (enables TSO/USO) and listen.udp_offloads (enables UDP GSO/GRO) in config, reducing CPU overhead at high throughput; both are off by default but will become the default in a future release.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
tun: use_offloads: true listen: udp_offloads: true
When routines is set above 1 on Linux, each I/O routine's thread can be pinned to a CPU using tun.pin_threads (default true), tun.pin_threads_key (pid or port), and tun.cpu_affinity, reducing latency jitter on multi-core hosts.
routines: 4 tun: pin_threads: true pin_threads_key: pid
CGO-free FIPSFIPSA set of US federal standards published by NIST that define approved cryptographic algorithms and modules; cyber tools use it to meet compliance requirements in government and regulated environments. 140-3 binaries can be built with make fips140 or make release-fips140; these builds restrict operation to P256 certificates and AES, and nebula-cert ca and keygen default to -curve P256.
Library embedders must update to new signatures: config.C.RegisterReloadCallback callbacks now take (*config.C, *config.Reload); ReloadConfig and ReloadConfigString return a config.ReloadResult (previously ReloadConfigString returned an error, now available as the result's Err); overlay.Device no longer embeds io.ReadWriteCloser; NewMultiQueueReader and SupportsMultiqueue are replaced by Queues; udp.Conn gained WriteBatch and IPv4Only; ListenOut takes a flush callback; ReloadConfig takes a *config.Reload; udp.NewListener and udp.NewGenericListener take a udp.Settings; SSH debug commands moved to a new commands package, sshd.RegisterCommand and sshd.Command are removed, sshd.NewSSHServer takes a *commands.Registry, and firewall.ProtoTCP, ProtoUDP, ProtoICMP, ProtoICMPv6 are now iputil.IPProtocol*.
Every SSH debug command is now reachable through a new local control socket using nebula ctl <command>; access is restricted to the running user, root, or a group named in local_api.control.group, and the socket path can be set per-instance with local_api.control.socket.
$ nebula ctl print-cert
Firewall rules with a group or groups field that YAML reads as a float, date, map, or nested array are now rejected at startup and reload instead of being silently coerced to a string (previously group: 1.10 matched group 1.1), and a null group/groups is rejected when no host, other group, or cidr limits the rule. A pki.blocklist entry that isn't a hex 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. fingerprint now causes Nebula to refuse to start or reload. On Linux, Nebula now refuses a tun.dev that already exists (including persistent devices from ip tuntap add or systemd-networkd). tun.mtu and any route mtu above 8937 is capped to 8937 with a warning, and nebula/nebula-service now refuse to start when given an argument that isn't a flag.
An opt-in info socket exposes node identity over HTTP: GET /v1/identity returns the local node's identity, GET /v1/identity/<vpn addr> returns another host's, and appending /pem to either path returns the certificate.
SSH debug server commands now exit with status 1 when they refuse, and report refusal as {"ok":false,"reason":...} when run with the newly-supported -json or -pretty flags (now accepted by every command). The start-cpu-profile, stop-cpu-profile, save-heap-profile, save-mutex-profile, and mutex-profile-fraction commands, along with sshd.sandbox_dir, have been removed.
A new experimental dns stanza lets a node override its host resolver, pointing dns.nameservers at queries for dns.match_domains or all lookups; it's off by default and supported only on macOS and Windows.
Nebula now batches outbound UDP sends on Linux, issuing up to 128 packets per sendmmsg call to improve throughput.
Building Nebula or embeddingembeddingLists of numbers a model produces to represent the meaning of text, images or code, so similar items sit close together. They power semantic search and RAG. it as a library now requires Go v1.27.3.
group that YAML reads as a float, a date, a map, or a nested array is now rejected on startup and reload instead of being silently converted to a string (e.g. group: 1.10 used to match group 1.1). A null group or groups is rejected when no host, other group, or cidr limits the rule.pki.blocklist entry that is not a hex SHA-256 fingerprint now causes Nebula to refuse to start and fail a reload.tun.dev that already exists (including a persistent device created with ip tuntap add or systemd-networkd); remove it and let Nebula create the device.tun.mtu and any route mtu above 8937 is capped to 8937 with a warning.-json or -pretty they report refusal as {"ok":false,"reason":...}. Every command now accepts -json and -pretty.nebula and nebula-service refuse to start when given an argument that is not a flag (previously ignored).config.C.RegisterReloadCallback callbacks now take (*config.C, *config.Reload); ReloadConfig and ReloadConfigString return a config.ReloadResult; ReloadConfigString used to return an error, now it is the result's Err.overlay.Device no longer embeds io.ReadWriteCloser; NewMultiQueueReader and SupportsMultiqueue are replaced by Queues; udp.Conn gained WriteBatch, IPv4Only; ListenOut takes a flush callback; ReloadConfig takes a *config.Reload; udp.NewListener and udp.NewGenericListener take a udp.Settings.commands package; sshd.RegisterCommand and sshd.Command are removed; sshd.NewSSHServer takes a *commands.Registry; firewall.ProtoTCP, ProtoUDP, ProtoICMP, and ProtoICMPv6 are now iputil.IPProtocol*.start-cpu-profile, stop-cpu-profile, save-heap-profile, save-mutex-profile, and mutex-profile-fraction commands, and sshd.sandbox_dir, are removed.Replace your VPN with Firezone, an open-source zero trust access platform built on WireGuard®. Connect users to anything, anywhere.
Firezone shipped performance improvements to WireGuardWireGuardA modern VPN protocol and Linux kernel module, maintained as an open-source project, that establishes encrypted tunnels using minimal code and fast cryptography, giving cyber tools a lightweight, high-performance transport layer. tunnel throughput and route handling, plus a redesigned Settings screen for choosing between US and EU control planes in the Linux GUIGUIA graphical user interface, controls such as windows, buttons and menus. It lets people use software visually rather than entering commands..︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
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.
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.
A redesigned Settings screen in the Firezone Linux GUI lets users switch between the US or EU control plane, useful for organizations that need data to stay in European infrastructure.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
Firezone Gateway and the Linux GUI now parallelize WireGuard encryption and decryption across worker threads, increasing tunnel throughput for established connections.
Packet-routing performance in the Linux GUI now benefits from caching of route resolution for established flows.
Immuta expands subscription policy enforcement beyond its native data integrations with a new custom securables capability for external assets.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
Use Immuta to provision governed access to data.
Immuta expands subscription policy enforcement beyond its native data integrations with a new custom securables capability for external assets.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
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.
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.
Immuta now supports custom securables, letting you register external assets — dashboards, files, APIs, and SaaSSaaSSoftware delivered over the internet and run on the provider's infrastructure, so the buyer pays a subscription instead of installing or operating anything locally. applications — in Immuta and control access to them with subscription policies.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄ Your tooling then calls Immuta's APIs or webhooks to retrieve access decisions and enforce them in the target system.
Varonis is the leading data and AI security platform that proactively reduces risk, enforces policies, and stops active threats with our AI-native automation.
Varonis now supports Azure Managed Identity authentication for connecting managed databases, removing the need for manually managed service-account credentials.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
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.
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.
Varonis can connect to Azure managed databases using Azure Managed Identity authentication instead of manually managed service-account credentials.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄ Supported databases include Azure SQL Server, PostgreSQL, and MySQL.
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 introduced an AI Agent Kill Switch that lets admins instantly revoke a single user's 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. client session without disrupting other agents or users, with support for both targeted and bulk revocation plus API automation.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
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.
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.
Admins can revoke a specific MCP user session by navigating to Administration > Identity Providers > Sessions tab and selecting Revoke session, scoped to one user and one MCP client without affecting other sessions.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄ Sessions in this view can also be filtered by user email, MCP client, or both and bulk-revoked using Revoke filtered, covering all sessions for one user or one MCP client. Revocation can be automated in incident-response workflows via the Aembit Cloud API. Sessions appear for MCP clients that connect through a Client-to-Gateway Access Policy using an OIDC ID Token Credential Provider with refresh tokenrefresh tokenA credential used to obtain a new access token without signing in again. It is sent to the authorization service rather than used directly to call the protected API. support enabled, and admins need Read/Write permission on Identity Providers to revoke.
Runlayer is an MCP gateway that enforces identity-mapped permissions and detects threats across the agents and MCP servers employees use.
Runlayer 2.0.0 hardens client-IP sourcing and session isolation to close cross-session manipulation and IP-spoofing gaps, while 1.41–1.42 add Terraform-managed skills with risk-policy enforcement, richer redacted session timelines, a compact tool-discovery header, and new Outlook, Google Docs, and LLMLLMA large language model, a neural network trained on large amounts of text to recognize patterns and generate language. Applications use it for tasks such as answering, summarizing and writing code.-budget-alert capabilities.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
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.
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.
Runlayer now determines client IP from the address your own proxy appended to X-Forwarded-For, never an earlier caller-supplied entry, so rate limits, ip_address on audit events, and the meta.request.ip policy field reflect one address per office or fleet rather than per workstation.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄ Self-hosted deployments with more than one appending proxy must set PRE_AUTH_TRUSTED_PROXY_HOPS (Terraform: trusted_proxy_hops) to the number of proxies, and CLIENT_IP_HEADER is now honoured only when it carries a single IP address.
PRE_AUTH_TRUSTED_PROXY_HOPS=2
The operator no longer injects AWS_ACCESS_KEY_ID, AWS_SECRET_ACCESS_KEY, or AWS_CA_BUNDLE on 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. pods; ExtendDB SigV4SigV4AWS's request-signing protocol that authenticates HTTP calls by hashing the request body, headers, and credentials into a signature; cyber tools use it to talk to AWS APIs without embedding long-lived secrets. now uses AWS profile extenddb via EXTENDDB_AWS_PROFILE (credentials under /etc/extenddb-aws, CA on the profile), and custom enable_db images must build the DynamoDB client from that profile when EXTENDDB_AWS_PROFILE is set. The Kubernetes operator also now rejects MCP servers that combine enable_db with an IRSAIRSAAn AWS mechanism that lets a Kubernetes pod assume an IAM role via a service account, so tools running in EKS can call AWS APIs without embedding long-lived credentials. annotation (eks.amazonaws.com/role-arn) on the MCP ServiceAccount, setting ExtendDBReady=False / IRSAConflict and leaving ExtendDB AWS env off the pod.
Redacted sessions and agent runs now preserve more data instead of hiding it: the API and the get_session MCP tool return the full event timeline (times, event types, tool names, model, tokens) with ***REDACTED*** in place of prompts, responses, tool input/output, and page URLs; the list_sessions and search_session_events MCP tools no longer include the first prompt or generated title in list and search rows; and get_agent_run_trace keeps its steps with placeholder text instead of hiding the timeline entirely.
Send X-Runlayer-Tool-Exposure: force-initial from any client to keep the Runlayer Plugin on search_tools / execute_tool instead of listing every connector tool, controlling compact tool discovery in agent harnessesagent harnessesThe software around an AI agent that manages its tools, execution loop, state and limits. It determines how the model's proposed actions are run and checked..
$ curl -H 'X-Runlayer-Tool-Exposure: force-initial' https://<your-runlayer-host>/mcp
Skill risk policy now runs on API writes: creating or updating a skill's files via the API or Terraform is rejected when the skill risk policy blocks the content, with nothing saved when a write is blocked; the new runlayer_skill Terraform provider resource manages skills and their complete file sets as code, with changes made in the app appearing as a plan diff and a policy-blocked file set failing the apply.
resource "runlayer_skill" "data_lookup" {
name = "data-lookup"
namespace = "finance"
path = "skills/data-lookup"
files = {
"skill.yaml" = file("skills/data-lookup/skill.yaml")
"handler.py" = file("skills/data-lookup/handler.py")
}
}
Set email alerts for LLM budget thresholds under Settings → LLM Gateway → Usage rules → Email alerts by configuring percentages; alerts fire once per percentage per budget window to organization settings administrators.
The Terraform provider now marks manual_oauth_client_secret and broker_vendor_client_secret on runlayer_server as sensitive, hiding them in plan and apply output; module outputs that reference them now require sensitive = true or the apply fails.
A tool call carrying someone else's session ID — via hook payload or the x-runlayer-session-id header — can no longer block that session, inject scan results into it, add security events to it, or mark an operator Stop as having taken effect in that session's history; hook events naming another session are no longer recorded into it.
Runlayer now drops X-Runlayer-User-Email, X-Runlayer-Identity-Token, and other identity-forward headers when a client sends them, so an upstream MCP server only ever sees the ones Runlayer injects; other custom headers such as a correlation ID still pass through.
ToolGuard score API responses now include a model_version field identifying the model version expected to score the request; include it when reporting scan results to Runlayer.
The Microsoft Outlook connector now supports creating mail folders, moving folders, and moving messages between folders, and the Google Docs connector now supports copying existing templates with native formatting preserved, inspecting document structure and styles, and attaching comments to specific text ranges; redeploy existing Outlook and Google Docs connectors to get the new tools.
Scheduled agent runs set up from an iMessage thread now post each run's result back via iMessage, and those set up from a Slack thread now post results back in that thread.
AWS_ACCESS_KEY_ID, AWS_SECRET_ACCESS_KEY, or AWS_CA_BUNDLE on MCP pods; ExtendDB SigV4 now uses AWS profile extenddb via EXTENDDB_AWS_PROFILE (credentials under /etc/extenddb-aws, CA on the profile). Custom enable_db images must build the DynamoDB client from that profile when EXTENDDB_AWS_PROFILE is set.ip_address on audit events, and the meta.request.ip policy field now use the address your own proxy appended to X-Forwarded-For, never an earlier caller-supplied entry. IP-range policies and per-IP rate limits (including device enrollment) now see one address per office or fleet rather than per workstation.PRE_AUTH_TRUSTED_PROXY_HOPS (Terraform: trusted_proxy_hops) to match; CLIENT_IP_HEADER is honoured only when it carries a single IP address.manual_oauth_client_secret and broker_vendor_client_secret on runlayer_server are now marked sensitive in Terraform; module outputs that reference them must add sensitive = true or the apply will fail.enable_db with an IRSA annotation (eks.amazonaws.com/role-arn) on the MCP ServiceAccount are now rejected by the operator (ExtendDBReady=False / IRSAConflict); remove one or the other to restore functionality.An ops AI Agent that understands your infrastructure, finds the root cause, and fixes it — right from Slack, Telegram, Lark or DingTalk.
ongrid's v0.19.0 release adds scoped external 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. access with durable operations, surfaces stream client runtime state through the Channels API and UI, and introduces pagination for large switch interface queries in the network view.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
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.
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.
The Channels API and UI now surface stream client runtime state, giving visibility into the live status of stream clients.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
ongrid now supports scoped external MCP access paired with durable operations, letting the agent interact with external MCP servers under constrained permissions while operations persist reliably.
ongrid now paginates large switch interface queries in the network view, improving handling of queries that return many interfaces.
h5i v0.5.2 ships four plugin-based capabilities: Chromium browser session capture with request auditing, a websec workbench for replaying requests, a recon plugin for endpoint discovery, and CI-gated security regression testing against OpenAPIOpenAPIA vendor-neutral specification for describing REST APIs in a machine-readable format, maintained by the OpenAPI Initiative. Cyber tools use it to auto-generate clients, documentation, and validation from a single source of truth. coverage.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
Build, hack, and verify web apps. Find vulnerabilities through red-teaming.
h5i v0.5.2 ships four plugin-based capabilities: Chromium browser session capture with request auditing, a websec workbench for replaying requests, a recon plugin for endpoint discovery, and CI-gated security regression testing against OpenAPIOpenAPIA vendor-neutral specification for describing REST APIs in a machine-readable format, maintained by the OpenAPI Initiative. Cyber tools use it to auto-generate clients, documentation, and validation from a single source of truth. coverage.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
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.
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.
h5i browser open, h5i browser snapshot, h5i browser requests, and h5i browser audit capture, policy-check, and log every HTTP request made by a Chromium session.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄ An allowlist is checked on every request and redirect hop, and --in adds an out-of-browser witness to the log.
$ h5i browser open https://staging.example.com
# interact with the app, then:
h5i browser requests
h5i browser audit
Install the recon plugin with h5i plugin install recon --from ./h5i-recon to record endpoints, parameters, and candidate URLs a target exposes, grounded in stored messages.
$ curl -L https://github.com/h5i-dev/h5i/releases/download/v0.5.2/h5i-recon-v0.5.2-x86_64-unknown-linux-musl.tar.gz | tar -xz h5i plugin install recon --from ./h5i-recon
Run repository-owned attack flows against a running server with h5i test .h5i-tests/tests --target http://localhost:3000 --min-coverage 80 to fail CI if OpenAPI coverage drops below the given percentage.
$ h5i test .h5i-tests/tests --target http://localhost:3000 --min-coverage 80
Install the websec workbench with h5i plugin install websec --from ./h5i-websec to replay and manipulate HTTP requests through the same policy and budget as the browser.
nono's v0.81.0 release adds a new policy-diagnostic command for Unix sockets, opt-in Linux namespaceLinux namespaceA Linux kernel feature that partitions global system resources—processes, network interfaces, mounts, users—so each namespace gets its own isolated view, letting tools sandbox environments without a full VM. isolation, a new runtime-state policy group, LandlockLandlockA Linux kernel access-control mechanism that lets an unprivileged process restrict its own filesystem and network access, giving cyber tools a way to sandbox themselves without root privileges. V9 detection, structured proxy denial logging, and a breaking change requiring fully qualified profile names.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
The nono runtime runs AI agents in isolated environments with policy-controlled filesystem, network, and credential access.
nono's v0.81.0 release adds a new policy-diagnostic command for Unix sockets, opt-in Linux namespaceLinux namespaceA Linux kernel feature that partitions global system resources—processes, network interfaces, mounts, users—so each namespace gets its own isolated view, letting tools sandbox environments without a full VM. isolation, a new runtime-state policy group, LandlockLandlockA Linux kernel access-control mechanism that lets an unprivileged process restrict its own filesystem and network access, giving cyber tools a way to sandbox themselves without root privileges. V9 detection, structured proxy denial logging, and a breaking change requiring fully qualified profile names.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
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.
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.
Query pathname Unix socket connect access with nono why --unix-socket <path> to diagnose AF_UNIX policy decisions, e.g.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄ checking whether a sandboxed agent can connect to /run/docker.sock.
$ nono why --unix-socket /run/docker.sock
Proxy deny log lines now include denial_category, giving operators a structured field to filter and alert on blocked network requests.
Profiles now require fully qualified package profile names, making profile references unambiguous across registries; bare/unqualified profile name references that worked in v0.80.0 will break on upgrade.
nono now detects Landlock ABI up to V9, extending future-proofing for Linux kernel upgrades while keeping enforcement at V6.
The exec gate on Linux now reads only the ELFELFThe standard binary format for executables, shared libraries, and object files on Linux and most Unix-like systems, giving cyber tools a predictable structure to parse, instrument, or analyze compiled code. regions it actually parses, reducing per-execution overhead and enabling lower-latency sandboxingsandboxingRunning code inside a restricted environment that limits what it can access or change. The isolation reduces the impact of mistakes or untrusted code, depending on the sandbox's controls. for workloads that spawn many processes.
Add the rtk_state policy group to grant runtime-state access in profiles.
Enable opt-in namespace isolation on Linux (NEP-0004) for stronger per-tool containment.