PAGES
Free Head is the only free newsletter you can have delivered to your inbox — the week’s best releases, every Friday. Sign up here →

This Week's Highlights — issue 014, October 2, 2026

THE CYBER TOOLCHAIN NO. 014
Head
THE WEEK'S TOP PICKS
PUBLISHED OCTOBER 2, 2026 · FRIDAYS
EDITIONS tail grep head ls diff uniq

The best tooling updates that shipped this week.

The week in numbers · Sep 25–Oct 1

Changes per week, last 12 weeks · Sep 25–Oct 1

Changes we read in that week's dailies, including docs and API changes. A gap means no daily was published.

596
Jul 10
426
Jul 17
399
Jul 24
497
Jul 31
410
Aug 7
515
Aug 14
336
Aug 21
468
Aug 28
593
Sep 4
591
Sep 11
732
Sep 18
640
Sep 25

Category moves this month

Checking Plays…

Every feature release from the 1326 tools on our watchlist goes into the daily newsletter. Once a week we read the whole field side by side and grade each release from 1 to 5 on three things, weighted equally:

  • Depth. How complete is the best new capability?
  • Impact. How much does it change what you can do?
  • Novelty. Has any other tool done this before? We check every capability against the history of every tool we track. Features most tools already have earn nothing here. Being first earns the most.

A tool is judged on everything it shipped that week, so a project that releases daily gets credit for the sum — and still only takes one slot. We would rather run a short list than a padded one.

VIEW
ISSUE VIEW scroll
INFORMATION DENSITY everything
ISSUE WIDTH reading width
Do you prefer scroll?

The picks in full

01 Cribl changelog · Sep 29, 2026 Security Data Pipeline

Cribl is an observability pipeline platform that helps organizations collect, process, route, and transform machine data from any source to any destination.

// WHY IT MADE THE LIST Depth 5/5 Impact 4/5 Novelty 5/5

Cribl Detect launches a full SIEM built on the data-pipeline platform — 8,000+ detection rules, detection posture management showing coverage gaps and broken rules, and in-pipeline detection that generates signals while telemetry is in motion, priced on infrastructure rather than per-GB. A pipeline vendor entering the SIEM market lets its customers consolidate detection where their data already flows.

At CriblCon 26, Cribl launched two major platform additions: Cribl Detect, a full 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. built on its pipeline, and StreamAI, an enterprise AI gateway — alongside a new agentic development environment, broader agent/integration access via 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., automatic telemetry parsing, and a growing app marketplace.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

└──▷ WHAT SHIPPED · 6 FEATURES 2 OF 6 FEATURES most completely described first
what's the number?

Each feature carries 0–100 for how completely the vendor documented it — not how big or important the work is. A major capability described in eight words scores low, and that is the finding.

  • depth0–40 what it does and how it works or what changed
  • specificity0–30 names real surfaces — APIs, flags, formats, limits, numbers
  • actionability0–30 enough to go use it — a named endpoint, flag, or config key tops this; a UI path is a starting point

Code and config rank first by construction: a documented endpoint, flag or config key scores at the top of specificity and actionability, so it sorts above a feature described only in prose. Under 60 the notes go thin — everything below the “thinner coverage” line is thinner documentation, not smaller work. Hover any meter for that feature's three sub-scores.

01 01
Cribl Detect SIEM with detection and posture management Not checked yet new 87

Cribl Detect is a complete SIEM built on the Cribl platform covering detection, coverage and posture management, investigation, response, and threat intelligence, available now for Cribl.Cloud customers at up to 50% less cost than incumbent SIEMsSIEMsSecurity Information and Event Management platforms that collect, normalize, and correlate log data across an infrastructure so cyber tools can detect threats and trigger alerts from a central point., priced on infrastructure rather than per-GB ingestion.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠ It includes Detection Posture Management that continuously shows coverage gaps, broken rules, and missing telemetry required to detect specific attacks, ships with a catalog of more than 8,000 detection rules with correlation across signals to surface multi-stage attacks, and generates detection signals directly in the data pipeline for real-time identification while telemetry is still in motion.

Get started with Cribl Detect on Cribl.Cloud to immediately access SIEM detection, posture management, and 8,000+ detection rules without migrating existing telemetry pipelines.
📍In Cribl.Cloud, log in to your organization, navigate to the Cribl Detect tile, and select Get Started to activate detection on your existing telemetry streams.
— Pricing model, rule count, and get-started steps all named 2026-09-29-1d8f4bc7
02 00
StreamAI enterprise AI gateway with model routing Not checked yet new 60

StreamAI is an enterprise AI gateway with a built-in model router that routes each prompt to the right model based on SecIT Bench benchmarks, offering free inference for customers using automatic routing. It enforces spending budgets and cutoffs, redacts sensitive data in both prompts and responses, and records every model call as audit-ready telemetry.

— Mechanism and controls named but no setup steps given 2026-09-29-1d8f4bc7
thinner coverage below
03 00
Agentic development environment on Detect's runtime Not checked yet new 48

A new agentic development environment lets teams build apps, workflows, and agents using natural language, running on the same agentic runtime that powers Cribl Detect, with compute, scheduling, logging, and permissions included.

— Describes capability and shared runtime, no usage detail 2026-09-29-1d8f4bc7
04 02
Cribl App Framework grows to 150+ apps Not checked yet improved 40

The Cribl App Framework has grown to more than 150 apps from Cribl, partners, and customers, available in the Cribl Marketplace.

— Names count and marketplace location, no further mechanism 2026-09-29-1d8f4bc7
05 00
Platform opened to agents via MCP and SDKs Not checked yet new 35

Cribl opens the platform to agents via MCP and SDKs, adding 70 new integrations this year.

— Names MCP and integration count but no how-to 2026-09-29-1d8f4bc7
06 00
AI auto-parsing of incoming telemetry Not checked yet new 20

Cribl adds AI auto-parsing of telemetry so incoming data is structured and usable the moment it arrives.

— Describes outcome only, no mechanism or configuration 2026-09-29-1d8f4bc7

Cribl launches StreamAI (AI gateway with model routing) and Cribl Detect (full 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. with 8,000+ detection rules) at CriblCon 26.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

└──▷ HOW TO FIND IT
Get started with Cribl Detect on Cribl.Cloud to immediately access SIEM detection, posture management, and 8,000+ detection rules without migrating existing telemetry pipelines.
📍In Cribl.Cloud, log in to your organization, navigate to the Cribl Detect tile, and select Get Started to activate detection on your existing telemetry streams.
  • Adds Cribl Detect, a complete SIEM built on the Cribl platform with detection, coverage and posture management, investigation, response, and threat intelligence — available now for Cribl.Cloud customers at up to 50% less cost than incumbent SIEMsSIEMsSecurity Information and Event Management platforms that collect, normalize, and correlate log data across an infrastructure so cyber tools can detect threats and trigger alerts from a central point., priced on infrastructure rather than per-GB ingestion.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠ Not checked yet
  • Adds StreamAI, an enterprise AI gateway with a built-in model router that routes each prompt to the right model based on SecIT Bench benchmarks, with free inference for customers using automatic routing. Not checked yet
  • StreamAI enforces spending budgets and cutoffs, redacts sensitive data in both prompts and responses, and records every model call as audit-ready telemetry. Not checked yet
  • Cribl Detect includes Detection Posture Management, continuously showing coverage gaps, broken rules, and missing telemetry required to detect specific attacks. Not checked yet
  • Cribl Detect ships with a catalog of more than 8,000 detection rules covering a broad range of threats, with correlation across signals to surface multi-stage attacks. Not checked yet
+5 more show less
  • Cribl Detect generates detection signals directly in the data pipeline, enabling real-time threat identification while telemetry is still in motion. Not checked yet
  • Adds an agentic development environment that lets teams build apps, workflows, and agents using natural language, running on the same agentic runtime that powers Cribl Detect with compute, scheduling, logging, and permissions included. Not checked yet
  • Opens the platform to agents via MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface. and SDKs, adding 70 new integrations this year. Not checked yet
  • Adds AI auto-parsing of telemetry so incoming data is structured and usable the moment it arrives. Not checked yet
  • Grows the Cribl App Framework to 150+ apps from Cribl, partners, and customers, available in the Cribl Marketplace. Not checked yet
  • Adds Cribl Detect, a complete SIEM built on the Cribl platform with detection, coverage and posture management, investigation, response, and threat intelligence — available now for Cribl.Cloud customers at up to 50% less cost than incumbent SIEMsSIEMsSecurity Information and Event Management platforms that collect, normalize, and correlate log data across an infrastructure so cyber tools can detect threats and trigger alerts from a central point., priced on infrastructure rather than per-GB ingestion.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠ Not checked yet
  • Adds StreamAI, an enterprise AI gateway with a built-in model router that routes each prompt to the right model based on SecIT Bench benchmarks, with free inference for customers using automatic routing. Not checked yet
  • StreamAI enforces spending budgets and cutoffs, redacts sensitive data in both prompts and responses, and records every model call as audit-ready telemetry. Not checked yet
  • Cribl Detect includes Detection Posture Management, continuously showing coverage gaps, broken rules, and missing telemetry required to detect specific attacks. Not checked yet
  • Cribl Detect ships with a catalog of more than 8,000 detection rules covering a broad range of threats, with correlation across signals to surface multi-stage attacks. Not checked yet
+4 more show less
  • Cribl Detect generates detection signals directly in the data pipeline, enabling real-time threat identification while telemetry is still in motion. Not checked yet
  • Adds an agentic development environment that lets teams build apps, workflows, and agents using natural language, running on the same agentic runtime that powers Cribl Detect with compute, scheduling, logging, and permissions included. Not checked yet
  • Opens the platform to agents via MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface. and SDKs, adding 70 new integrations this year. Not checked yet
  • Adds AI auto-parsing of telemetry so incoming data is structured and usable the moment it arrives. Not checked yet

1 more in Everything

Good pick?
Good reason?
02 Qualys VMDR changelog · Sep 29, 2026 Vulnerability & Exposure Management

Qualys VMDR is a cloud-based vulnerability management platform that continuously discovers, assesses, prioritizes, and remediates security vulnerabilities across IT assets.

// WHY IT MADE THE LIST Depth 4/5 Impact 5/5 Novelty 5/5

InstaScan delivers a vulnerability finding within an hour of a vendor disclosure — closing the window where exposure is live but unassessed — and pairs it with prioritization that tests exploitability against the environment's actual firewall and EDR controls rather than theoretical CVSS. No tracked tool ships either capability.

Qualys VMDR introduced InstaScan for near-instant vulnerability detection following vendor disclosures, paired with hyper-prioritization that tests exploitability against an environment's actual security controls.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

└──▷ WHAT SHIPPED · 2 FEATURES 2 OF 2 FEATURES most completely described first
what's the number?

Each feature carries 0–100 for how completely the vendor documented it — not how big or important the work is. A major capability described in eight words scores low, and that is the finding.

  • depth0–40 what it does and how it works or what changed
  • specificity0–30 names real surfaces — APIs, flags, formats, limits, numbers
  • actionability0–30 enough to go use it — a named endpoint, flag, or config key tops this; a UI path is a starting point

Code and config rank first by construction: a documented endpoint, flag or config key scores at the top of specificity and actionability, so it sorts above a feature described only in prose. Under 60 the notes go thin — everything below the “thinner coverage” line is thinner documentation, not smaller work. Hover any meter for that feature's three sub-scores.

01 01
InstaScan scanless detection Not checked yet new 45

InstaScan is a scanless scanning capability that delivers a finding within an hour of a vendor disclosure, enabling much faster detection than traditional scan cycles.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

— Describes mechanism and timing but no config or command 2026-09-29-9128b856
02 02
Hyper-prioritization against live controls Not checked yet new 40

Hyper-prioritization tests exploitability against a given environment's actual firewall and EDREDREndpoint detection and response, monitoring computers for suspicious behavior and providing investigation and response controls, such as isolating a device. controls, rather than relying on theoretical scoring, to better prioritize remediation.

— Explains approach but no concrete settings or workflow steps 2026-09-29-9128b856

Qualys VMDR adds InstaScan for sub-hour detection from vendor disclosure, with hyper-prioritization against live controls.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

  • Introduces InstaScan, a scanless scanning capability that delivers a finding within an hour of a vendor disclosure.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠ Not checked yet
  • Adds hyper-prioritization that tests exploitability against a given environment's actual firewall and EDREDREndpoint detection and response, monitoring computers for suspicious behavior and providing investigation and response controls, such as isolating a device. controls, rather than using theoretical scoring. Not checked yet

2 details in Everything

Good pick?
Good reason?
03 Qualys Patch Management changelog · Sep 29, 2026 Vulnerability & Exposure Management

Qualys Patch Management is a cloud-based tool that automates the discovery, prioritization, and deployment of patches across IT assets to reduce vulnerabilities.

// WHY IT MADE THE LIST Depth 4/5 Impact 4/5 Novelty 4/5

Patch Blocking Rules let you define a bad patch by QQL query and block it across every deployment job at once, and AI-Powered Patch Reliability Scoring rates each patch from real-world internet feedback plus Qualys rollback and vuln-reopen telemetry — automatically skipping low-reliability patches from zero-touch jobs so a known-bad update never reaches production.

Qualys TruRiskTruRiskA proprietary risk-scoring methodology from Qualys that translates vulnerability and asset data into a numeric business-risk score, giving cyber tools a single prioritization signal instead of raw CVSS values. Eliminate adds guardrails to automated patching: blocking rules to halt bad patches fleet-wide, AI-driven reliability scoring to keep low-confidence patches out of zero-touch jobs, and curated mitigations to cover the gap while patches are held.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

└──▷ WHAT SHIPPED · 3 FEATURES 2 OF 3 FEATURES most completely described first
what's the number?

Each feature carries 0–100 for how completely the vendor documented it — not how big or important the work is. A major capability described in eight words scores low, and that is the finding.

  • depth0–40 what it does and how it works or what changed
  • specificity0–30 names real surfaces — APIs, flags, formats, limits, numbers
  • actionability0–30 enough to go use it — a named endpoint, flag, or config key tops this; a UI path is a starting point

Code and config rank first by construction: a documented endpoint, flag or config key scores at the top of specificity and actionability, so it sorts above a feature described only in prose. Under 60 the notes go thin — everything below the “thinner coverage” line is thinner documentation, not smaller work. Hover any meter for that feature's three sub-scores.

01 01
Patch Blocking Rules in TruRisk Eliminate Not checked yet new 88

A new Patch Blocking Rules page under TE > Configuration > Patch Blocking Rules (Windows, Linux, and Mac support added in release 4.2) lets admins define patches to block via a QQL query (e.g.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠ patch.kb: [KB5002907, KB5124008]) and block them across every deployment job simultaneously, without editing individual jobs. Scope can be restricted to up to five asset tags per rule, up to 10 rules are allowed per platform, and each rule requires a name and a blocking reason.

Immediately stop a vendor-paused or known-bad Windows patch from deploying across your entire environment without touching individual patch jobs.
📍In TruRisk Eliminate, go to TE › Configuration › Patch Blocking Rules, select Windows, click Create Rule. Enter a name (e.g. Block paused Sep-2026 updates) and a blocking reason. Under Criteria (QQL), enter: patch.kb: [KB5002907, KB5124008, KB5124012, KB5123099] and click Preview to confirm the match. Leave asset tags empty to cover all assets, then save.
Protect production assets while a test group validates a new patch by scoping a blocking rule to production asset tags.
📍In TruRisk Eliminate, go to TE › Configuration › Patch Blocking Rules, select Windows, click Create Rule. Enter a name (e.g. Hold KB5002907 from prod during validation) and a blocking reason. Under Criteria (QQL), enter: patch.kb: [KB5002907] and click Preview. Under asset scope, add up to five production asset tags to restrict the block to production only, then save.
— Names exact menu path, QQL syntax, and limits 2026-09-29-7aca09ff
02 02
AI-Powered Patch Reliability Scoring Not checked yet new 70

Combines 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. analysis of real-world internet feedback with Qualys rollback and vulnerability-reopen telemetry to rate each patch as High, Medium, Low, or Unidentified. When reliability intelligence is added to a zero-touch job's QQL in the Select Patches step, Low-reliability patches are automatically skipped from deployment while others continue.

Keep zero-touch patch jobs running while automatically skipping low-reliability patches by adding reliability intelligence to the job's patch-selection query.
📍In the zero-touch patch job's Select Patches step, update the QQL to include reliability criteria alongside your existing filters so that Low-reliability patches — such as those currently rated Low by AI-Powered Patch Reliability Scoring — are excluded from automatic deployment while all other patches continue to deploy.
— Explains rating mechanism and where to apply it 2026-09-29-7aca09ff
thinner coverage below
03 00
Curated mitigations for blocked or risky patches Not checked yet new 23

Qualys-curated mitigations are introduced as a complementary layer to reduce exposure while a blocked or low-reliability patch is held in testing and staging.

— Described only in prose, no mechanism or surface named 2026-09-29-7aca09ff

Qualys TruRiskTruRiskA proprietary risk-scoring methodology from Qualys that translates vulnerability and asset data into a numeric business-risk score, giving cyber tools a single prioritization signal instead of raw CVSS values. Eliminate adds Patch Blocking Rules and AI-Powered Patch Reliability Scoring to keep bad patches out of automation.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

└──▷ HOW TO FIND IT
Immediately stop a vendor-paused or known-bad Windows patch from deploying across your entire environment without touching individual patch jobs.
📍In TruRisk Eliminate, go to TE › Configuration › Patch Blocking Rules, select Windows, click Create Rule. Enter a name (e.g. Block paused Sep-2026 updates) and a blocking reason. Under Criteria (QQL), enter: patch.kb: [KB5002907, KB5124008, KB5124012, KB5123099] and click Preview to confirm the match. Leave asset tags empty to cover all assets, then save.
Protect production assets while a test group validates a new patch by scoping a blocking rule to production asset tags.
📍In TruRisk Eliminate, go to TE › Configuration › Patch Blocking Rules, select Windows, click Create Rule. Enter a name (e.g. Hold KB5002907 from prod during validation) and a blocking reason. Under Criteria (QQL), enter: patch.kb: [KB5002907] and click Preview. Under asset scope, add up to five production asset tags to restrict the block to production only, then save.
Keep zero-touch patch jobs running while automatically skipping low-reliability patches by adding reliability intelligence to the job's patch-selection query.
📍In the zero-touch patch job's Select Patches step, update the QQL to include reliability criteria alongside your existing filters so that Low-reliability patches — such as those currently rated Low by AI-Powered Patch Reliability Scoring — are excluded from automatic deployment while all other patches continue to deploy.
  • Adds Patch Blocking Rules in TE > Configuration > Patch Blocking Rules (available for Windows, Linux, and Mac in release 4.2), letting you define patches via a QQL query and block them across every deployment job at once without editing individual jobs; scope is controlled by up to five asset tags per rule, with up to 10 rules per platform, and each rule requires a name and blocking reason.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠ Not checked yet
  • Adds AI-Powered Patch Reliability Scoring that combines 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. analysis of real-world internet feedback with Qualys rollback and vulnerability-reopen telemetry to rate each patch High, Medium, Low, or Unidentified — automatically skipping Low-reliability patches from zero-touch jobs when reliability intelligence is added to the job's QQL in the Select Patches step. Not checked yet
  • Introduces Qualys-curated mitigations as a complementary layer to reduce exposure while a blocked or low-reliability patch is held in testing and staging. Not checked yet

3 details in Everything

Good pick?
Good reason?
04 Metasploit Pro October 1, 2026: 5.1.0-2026093001 Exploitation & C2

Metasploit Pro is a commercial penetration testing framework for identifying, exploiting, and validating security vulnerabilities in systems.

// WHY IT MADE THE LIST Depth 4/5 Impact 4/5 Novelty 4/5

Native Kerberos authentication relay lands in the framework's relay stack, including the esc8_kerberos module for AD CS Web Enrollment and IPv6/DHCPv6 DNS-takeover coercion modules to drive the chain — alongside OPC-UA enumeration and Windows AArch64 payloads across 41 new modules.

Metasploit Pro 5.1.0 ships 41 new modules headlined by native KerberosKerberosA network authentication protocol that uses tickets issued by a trusted service. Users and services can prove their identities without sending a password to each application. authentication relay (with ESC8ESC8An Active Directory privilege escalation path where an attacker abuses the Web Enrollment interface of a Microsoft Certificate Authority to coerce NTLM authentication and relay it to obtain a certificate for a privileged account. exploitation and IPv6IPv6The current version of the Internet Protocol, standardized by the IETF, using 128-bit addresses to replace the exhausted IPv4 space. Cyber tools target it because modern networks run dual-stack or IPv6-only. DNS-takeover coercion), new OPC-UAOPC-UAAn open, platform-neutral communication protocol for industrial automation systems, maintained by the OPC Foundation, giving cyber tools a standardized way to read and monitor data from PLCs, sensors, and SCADA equipment. and WinRM scanning capabilities, Windows AArch64AArch64The 64-bit execution state of the ARM architecture, defined by Arm Ltd., offering a wider register file and cleaner instruction set than 32-bit ARM; cyber tools target it to run natively on ARM-based servers, phones, and embedded devices. and MIPS payload support, and now requires authentication on JSON RPC by default.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

└──▷ WHAT SHIPPED · 12 FEATURES 1 OF 12 FEATURES most completely described first
what's the number?

Each feature carries 0–100 for how completely the vendor documented it — not how big or important the work is. A major capability described in eight words scores low, and that is the finding.

  • depth0–40 what it does and how it works or what changed
  • specificity0–30 names real surfaces — APIs, flags, formats, limits, numbers
  • actionability0–30 enough to go use it — a named endpoint, flag, or config key tops this; a UI path is a starting point

Code and config rank first by construction: a documented endpoint, flag or config key scores at the top of specificity and actionability, so it sorts above a feature described only in prose. Under 60 the notes go thin — everything below the “thinner coverage” line is thinner documentation, not smaller work. Hover any meter for that feature's three sub-scores.

01 00
WinRM login session type selection Not checked yet improved 85

Adds a SessionType option to auxiliary/scanner/winrm/winrm_login to create a cmd-backed shell, a PSRP-backed PowerShell session, or automatically fall back to PowerShell when cmd shell creation is denied.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

On a WinRMWinRMA Windows-native remote management protocol from Microsoft, implementing the WS-Management standard, that lets tools execute commands and access data on remote Windows hosts over HTTP or HTTPS. login attempt, automatically fall back to a PowerShell session when cmd shell creation is denied — useful against hardened Windows targets.
$ use auxiliary/scanner/winrm/winrm_login
set RHOSTS 192.168.1.50
set USERNAME administrator
set PASSWORD P@ssw0rd
set SessionType powershell
run
— Exact option and module path plus a runnable fallback example October 1, 2026: 5.1.0-2026093001
02 01
Kerberos authentication relay stack Not checked yet new 80

The relay stack gains native Kerberos authentication relay capabilities, including the new esc8_kerberos auxiliary module for exploiting AD CSAD CSActive Directory Certificate Services, Microsoft's PKI role for issuing and managing X.509 certificates inside a Windows domain; cyber tools target it because misconfigurations frequently allow privilege escalation or domain compromise. Web Enrollment (ESC8), auxiliary/spoof/ipv6/ipv6_ra_dns_takeover and auxiliary/spoof/dhcp/dhcpv6_dns_takeover for native IPv6 DNS-takeover coercion as part of a Kerberos relay attack chain, KerberosTicketTrace support added to auxiliary/admin/kerberos/forge_ticket, and a new shared low-level implementation Rex::Proto::Gss::KerberosToken for handling Kerberos tokens carried by GSS-API and SPNEGO.

— Names every module and class but gives no usage example October 1, 2026: 5.1.0-2026093001
03 00
OPC-UA server enumeration scanner Not checked yet new 80

Adds auxiliary/scanner/scada/opcua_enum, a scanner module that detects OPC-UA servers speaking the OPC-UA TCPTCPA network protocol that delivers a reliable, ordered stream of bytes between two endpoints. It detects lost data and retransmits it. binary transport (opc.tcp://), useful for enumerating exposed industrial control endpoints.

Scan an 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. network segment for OPC-UA servers exposing the binary TCP transport to enumerate exposed industrial control endpoints.
$ use auxiliary/scanner/scada/opcua_enum
set RHOSTS 10.10.20.0/24
run
— Module path, protocol scheme, and runnable scan command given October 1, 2026: 5.1.0-2026093001
04 00
SPIP blind SQL injection scanner Not checked yet new 65

Adds auxiliary/scanner/http/spip_annee_sqli, which exploits a blind SQL injectionSQL injectionAn attack that makes an application treat untrusted input as part of a database command. It can expose or modify data by changing what the query does. in SPIP's date column escaping logic.

— Exact module path and vulnerability mechanism named, no example run October 1, 2026: 5.1.0-2026093001
05 00
Windows AArch64 payload support Not checked yet new 60

Adds Windows AArch64 staged shell payloads and extends windows/smb/psexec to work with AArch64 targets.

— Names the module and architecture, no usage walkthrough October 1, 2026: 5.1.0-2026093001
06 00
ZoomEye search module v2 API support Not checked yet improved 60

Updates the ZoomEye search module to support the ZoomEye v2 API, adds a three-attempt retry limit, and caps results per page at 10.

— Concrete limits named but no sample invocation October 1, 2026: 5.1.0-2026093001
thinner coverage below
07 00
JSON RPC requires authentication by default Not checked yet breaking 55

JSON RPCRPCRemote Procedure Call, a protocol that lets a program execute a function on another machine as if it were local, giving cyber tools a structured way to communicate across processes or hosts without managing raw sockets. support now requires authentication by default, either via the database with user credentials or an auth token; existing unauthenticated JSON RPC clients will be rejected on upgrade.

— States the new requirement but no exact config key or migration step October 1, 2026: 5.1.0-2026093001
08 00
MIPS payload additions and updates Not checked yet improved 45

Adds a MIPS64 exec payload, updates the MIPSEL and MIPSBE exec payloads, and adds prepends for all three architectures.

— Names architectures but no usage context given October 1, 2026: 5.1.0-2026093001
09 00
msfconsole usability and startup improvements Not checked yet improved 45

msfconsole adds MSFLOGO=nyan-cat.txt environment variable support for selecting the nyan cat banner, a new command-line flag to print the count of modules, and improved start-up performance.

— Env var named exactly, but the new flag's name is not given October 1, 2026: 5.1.0-2026093001
10 00
Persistence module updates Not checked yet improved 45

Adds a new multi-platform persistence module for Kate, and updates existing persistence modules to use create_process instead of cmd_exec.

— Names the mechanism swap but little on the new Kate module October 1, 2026: 5.1.0-2026093001
11 00
New reconnaissance and evasion modules Not checked yet new 40

Adds a scanner module that enumerates ports on a host to determine whether each is a Metasploit reverse handler and what kind of shell it expects, an auxiliary scanner that fingerprints Inductive Automation Ignition gateways across all major version families by probing unauthenticated info endpoints, and a Linux x64 sandbox-evasion module that performs runtime environment checks and aborts execution when a likely sandbox or VM is detected.

— Mechanisms described but no module paths given October 1, 2026: 5.1.0-2026093001
12 00
Windows session writable check in post mixin Not checked yet new 30

Adds support to the post mixin for Windows sessions to run the .writable? method.

— Bare mention of the method with no usage detail October 1, 2026: 5.1.0-2026093001
└──▷ BREAKING ON UPGRADE
  • !
    JSON RPC now requires authentication by default (via database credentials or auth token); existing unauthenticated JSON RPC clients will be rejected on upgrade.

Metasploit Pro 5.1.0-2026093001 adds 41 new modules including OPC-UAOPC-UAAn open, platform-neutral communication protocol for industrial automation systems, maintained by the OPC Foundation, giving cyber tools a standardized way to read and monitor data from PLCs, sensors, and SCADA equipment., KerberosKerberosA network authentication protocol that uses tickets issued by a trusted service. Users and services can prove their identities without sending a password to each application. relay, IPv6IPv6The current version of the Internet Protocol, standardized by the IETF, using 128-bit addresses to replace the exhausted IPv4 space. Cyber tools target it because modern networks run dual-stack or IPv6-only. DNS-takeover, and Windows AArch64AArch64The 64-bit execution state of the ARM architecture, defined by Arm Ltd., offering a wider register file and cleaner instruction set than 32-bit ARM; cyber tools target it to run natively on ARM-based servers, phones, and embedded devices. payloads.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

└──▷ TRY IT
Scan an 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. network segment for OPC-UA servers exposing the binary TCPTCPA network protocol that delivers a reliable, ordered stream of bytes between two endpoints. It detects lost data and retransmits it. transport to enumerate exposed industrial control endpoints.
$ use auxiliary/scanner/scada/opcua_enum
set RHOSTS 10.10.20.0/24
run
On a WinRMWinRMA Windows-native remote management protocol from Microsoft, implementing the WS-Management standard, that lets tools execute commands and access data on remote Windows hosts over HTTP or HTTPS. login attempt, automatically fall back to a PowerShell session when cmd shell creation is denied — useful against hardened Windows targets.
$ use auxiliary/scanner/winrm/winrm_login
set RHOSTS 192.168.1.50
set USERNAME administrator
set PASSWORD P@ssw0rd
set SessionType powershell
run
  • Adds auxiliary/scanner/scada/opcua_enum, a scanner module that detects OPC-UA servers speaking the OPC-UA TCP binary transport (opc.tcp://).⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠ Not checked yet
  • Adds SessionType option to auxiliary/scanner/winrm/winrm_login to create a cmd-backed shell, a PSRP-backed PowerShell session, or automatically fall back to PowerShell when cmd shell creation is denied. Not checked yet
  • Adds MSFLOGO=nyan-cat.txt environment variable support to msfconsole for selecting the nyan cat banner. Not checked yet
  • Adds a new command-line flag to msfconsole to print the count of modules. Not checked yet
  • Updates JSON RPCRPCRemote Procedure Call, a protocol that lets a program execute a function on another machine as if it were local, giving cyber tools a structured way to communicate across processes or hosts without managing raw sockets. support to now require authentication by default, either via the database with user credentials or an auth token. Not checked yet
+16 more show less
  • Introduces native Kerberos authentication relay capabilities to the framework's relay stack, including the new esc8_kerberos auxiliary module for exploiting AD CSAD CSActive Directory Certificate Services, Microsoft's PKI role for issuing and managing X.509 certificates inside a Windows domain; cyber tools target it because misconfigurations frequently allow privilege escalation or domain compromise. Web Enrollment (ESC8ESC8An Active Directory privilege escalation path where an attacker abuses the Web Enrollment interface of a Microsoft Certificate Authority to coerce NTLM authentication and relay it to obtain a certificate for a privileged account.). Not checked yet
  • Adds auxiliary/spoof/ipv6/ipv6_ra_dns_takeover and auxiliary/spoof/dhcp/dhcpv6_dns_takeover modules to enable native IPv6 DNS-takeover coercion as part of a Kerberos authentication relay attack chain. Not checked yet
  • Adds KerberosTicketTrace support to the auxiliary/admin/kerberos/forge_ticket module. Not checked yet
  • Introduces Rex::Proto::Gss::KerberosToken as a shared low-level implementation for handling Kerberos tokens carried by GSS-API and SPNEGO. Not checked yet
  • Adds Windows AArch64 staged shell payloads. Not checked yet
  • Extends windows/smb/psexec to work with AArch64 targets. Not checked yet
  • Adds a MIPS64 exec payload, updates the MIPSEL and MIPSBE exec payloads, and adds prepends for all three. Not checked yet
  • Adds a multi-platform persistence module for Kate. Not checked yet
  • Adds a scanner module that enumerates ports on a host to determine whether each is a Metasploit reverse handler and what kind of shell it expects. Not checked yet
  • Adds an auxiliary scanner module that fingerprints Inductive Automation Ignition gateways across all major version families by probing unauthenticated info endpoints. Not checked yet
  • Adds a Linux x64 sandbox-evasion module that performs runtime environment checks and aborts execution when a likely sandbox or VM is detected. Not checked yet
  • Adds auxiliary/scanner/http/spip_annee_sqli, which exploits a blind SQL injectionSQL injectionAn attack that makes an application treat untrusted input as part of a database command. It can expose or modify data by changing what the query does. in SPIP's date column escaping logic. Not checked yet
  • Updates the ZoomEye search module to support the ZoomEye v2 API, adds a three-attempt retry limit, and caps results per page at 10. Not checked yet
  • Adds support to the post mixin for Windows sessions to run the .writable? method. Not checked yet
  • Updates persistence modules to use create_process instead of cmd_exec. Not checked yet
  • Improves start-up performance of msfconsole. Not checked yet
  • Adds auxiliary/scanner/scada/opcua_enum, a scanner module that detects OPC-UA servers speaking the OPC-UA TCP binary transport (opc.tcp://).⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠ Not checked yet
  • Introduces native Kerberos authentication relay capabilities to the framework's relay stack, including the new esc8_kerberos auxiliary module for exploiting AD CSAD CSActive Directory Certificate Services, Microsoft's PKI role for issuing and managing X.509 certificates inside a Windows domain; cyber tools target it because misconfigurations frequently allow privilege escalation or domain compromise. Web Enrollment (ESC8ESC8An Active Directory privilege escalation path where an attacker abuses the Web Enrollment interface of a Microsoft Certificate Authority to coerce NTLM authentication and relay it to obtain a certificate for a privileged account.). Not checked yet
  • Adds auxiliary/spoof/ipv6/ipv6_ra_dns_takeover and auxiliary/spoof/dhcp/dhcpv6_dns_takeover modules to enable native IPv6 DNS-takeover coercion as part of a Kerberos authentication relay attack chain. Not checked yet
  • Introduces Rex::Proto::Gss::KerberosToken as a shared low-level implementation for handling Kerberos tokens carried by GSS-API and SPNEGO. Not checked yet
  • Adds Windows AArch64 staged shell payloads. Not checked yet
+5 more show less
  • Adds a multi-platform persistence module for Kate. Not checked yet
  • Adds a scanner module that enumerates ports on a host to determine whether each is a Metasploit reverse handler and what kind of shell it expects. Not checked yet
  • Adds an auxiliary scanner module that fingerprints Inductive Automation Ignition gateways across all major version families by probing unauthenticated info endpoints. Not checked yet
  • Adds a Linux x64 sandbox-evasion module that performs runtime environment checks and aborts execution when a likely sandbox or VM is detected. Not checked yet
  • Adds auxiliary/scanner/http/spip_annee_sqli, which exploits a blind SQL injectionSQL injectionAn attack that makes an application treat untrusted input as part of a database command. It can expose or modify data by changing what the query does. in SPIP's date column escaping logic. Not checked yet

11 more in Everything

└──▷ BREAKING ON UPGRADE
  • !JSON RPC now requires authentication by default (via database credentials or auth token); existing unauthenticated JSON RPC clients will be rejected on upgrade.
Good pick?
Good reason?
05 Bedrock Security announced · Sep 29, 2026 Data Security & DLP

More than DSPM: Bedrock's AI and data security platform unifies data, identity, and AI context so teams see real risk — not noise — without moving data.

// WHY IT MADE THE LIST Depth 4/5 Impact 4/5 Novelty 4/5

Agent DLP sits inline at the agent gateway via AWS AgentCore and LiteLLM hooks, inspecting every tool call in both directions to mask, redact, or block sensitive data in real time — closing the gap between declared policy and actual agent behavior, with an observe mode to validate coverage first.

Bedrock Security shipped a wave of AI governance capability: a new ArgusAI product (AI Data Bill of MaterialsBill of MaterialsA structured list of every component, library, and dependency that makes up a piece of software or hardware, used by cyber tools to track what is inside a product and surface known vulnerabilities in its parts., Guardrail Gap Analysis, natural-language policy and investigation), a graph-backed Metadata Lake with Copilot, AI Agents for DSPMDSPMData security posture management, locating sensitive data and checking how it is protected and who can reach it. It helps teams prioritize risks around their data stores. and a free Snowflake tier, an 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. exposing 50+ metadata signals to AI agents, and inline Agent DLPDLPData Loss Prevention: a category of controls that inspect and block sensitive data from leaving an organisation via email, web, or removable media, giving cyber tools a policy enforcement layer over data in motion and at rest. runtime enforcement extended from AWS AgentCore/LiteLLM to NVIDIA OpenShell.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

└──▷ WHAT SHIPPED · 11 FEATURES 3 OF 11 FEATURES most completely described first
what's the number?

Each feature carries 0–100 for how completely the vendor documented it — not how big or important the work is. A major capability described in eight words scores low, and that is the finding.

  • depth0–40 what it does and how it works or what changed
  • specificity0–30 names real surfaces — APIs, flags, formats, limits, numbers
  • actionability0–30 enough to go use it — a named endpoint, flag, or config key tops this; a UI path is a starting point

Code and config rank first by construction: a documented endpoint, flag or config key scores at the top of specificity and actionability, so it sorts above a feature described only in prose. Under 60 the notes go thin — everything below the “thinner coverage” line is thinner documentation, not smaller work. Hover any meter for that feature's three sub-scores.

01 01
Agent DLP extended to NVIDIA OpenShell Not checked yet new 77

Extends Agent DLP to NVIDIA OpenShell via OpenShell's Supervisor Middleware, evaluating each agent action against enterprise data policies and returning an allow/block decision enforced at the sandbox boundary.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠ Supports an observe-only mode to validate coverage before enforcing, exposes an agent's blast radius (what data its permissions can reach, what combinations it forms, where policy would be violated) at design time, and applies the same data policies across every invocation path an agent may use — MCP tools, command-line invocations, and agent-written code — from a single OpenShell control point. Restrictions travel across table writes, tool handoffs, and agent-to-agent transfers based on the Metadata Lake knowledge graphknowledge graphA representation of entities and the relationships between them. Software can follow those connections to answer questions or combine information from different sources. rather than payload pattern matching.

— Detailed mechanism and scope, no runnable command given launch-20260928-00b820b2
02 00
Natural Language Policy and Investigation in ArgusAI Not checked yet new 76

Adds Natural Language Policy and Investigation in ArgusAI, letting teams articulate data controls in plain English (e.g., "Only members of HR may view employment data" or "EU personal data must remain within EU regions") and have them translated into enforceable controls across connected systems. Extends the same natural-language interface to investigation over the DBOM and guardrail context, so queries like "Which models can access PHIPHIProtected Health Information: individually identifiable health data regulated under the US HIPAA law, whose strict handling requirements drive cyber tools to audit, mask, or restrict access to it. data?" or "Which identities interacted with customer 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. through an agent?" return answers with lineage and access evidence from the Metadata Lake.

Investigate which AI models in your Amazon Bedrock environment can access protected health information, using plain-English natural language queries.
📍In the Bedrock Data Platform, navigate to ArgusAI › Natural Language Investigation and enter: 'Which models can access PHI data?' — the system returns results with lineage and access evidence from the Metadata Lake.
Enforce a data residency policy across AI systems without writing control-plane code — useful for EU compliance requirements.
📍In the Bedrock Data Platform, go to Natural Language Policy and enter: 'EU personal data must remain within EU regions.' — Bedrock Data translates the invariant into enforceable controls across connected systems including AI agents and Knowledge Bases.
— Includes exact UI navigation and example queries to run launch-20260929-1fe46b0a
03 02
Agent DLP runtime enforcement in ArgusAI Not checked yet new 73

Adds Agent DLP to ArgusAI with native hooks for AWS AgentCore and LiteLLM, sitting inline at the agent gateway to inspect every tool call in both directions (outbound requests to MCP tools and inbound tool responses) without acting as a proxy. Enforces data access and regulatory policies to mask, redact, or block sensitive data exposure in real time, logs every decision with target, action, data types, and verdict (blocked, modified, redacted, or observed), and supports an observe mode for monitoring-only enforcement scenarios.

— Mechanism and verdict values named but no setup command launch-20260929-0b6583d7
04 00
MCP Server for the Metadata Lake Not checked yet new 65

Adds an MCP (Model Context ProtocolModel Context ProtocolModel 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 as a standardized gateway between AI agents and Bedrock's metadata lake, exposing data sensitivity, risk, lineage, access, and usage signals via the open MCP protocol. Surfaces over 50 distinct metadata signals as queryable context for agentic AI workflows and automated governance pipelines, letting AI agents and external applications — including legal and compliance teams — query the metadata lake without custom point-to-point integrations. Builds on the existing Bedrock Metadata Lake Copilot and audits AI model interactions with enterprise data for traceability and policy alignment.

— Protocol and signal count named, no endpoint or command shown launch-20260929-f953fb1a
thinner coverage below
05 03
Bedrock Metadata Lake Not checked yet new 58

Introduces the Bedrock Metadata Lake, a graph-backed centralized repository that continuously catalogs enterprise data stores across SaaSSaaSSoftware delivered over the internet and billed by subscription, hosted and operated by the vendor so the buyer skips installation and infrastructure management., PaaS, and IaaS, capturing location, sensitivity, entitlements, lineage, usage, risk, and more than fifty additional parameters.

— Clear scope and scale, no access steps given launch-20260929-d9dffb84
06 00
Bedrock Free for Snowflake Not checked yet new 55

Launches Bedrock Free for Snowflake, a free tier delivering automated sensitive-data discovery and classification for Snowflake environments, including access to Metadata Lake Copilot, with upgrade paths to entitlement chain and usage analysis and automated policy capabilities.

— Names tier and target platform, no setup steps given launch-20260929-d9dffb84
07 00
AI Agents for DSPM Not checked yet new 50

Introduces Bedrock AI Agents for DSPM, which proactively analyze the metadata lake for policy violations and new risks, gather context automatically, suggest next steps, and guide users through incident-response resolution workflows.

— Describes behavior but no concrete workflow steps launch-20260929-d9dffb84
08 00
Guardrail Gap Analysis in ArgusAI Not checked yet new 48

Adds Guardrail Gap Analysis in ArgusAI, which compares policy intent against model data exposure drawn from the Metadata Lake and generates targeted remediations to suppress or mask sensitive data at the source or agent runtime.

— Describes comparison logic but no concrete interface shown launch-20260929-1fe46b0a
09 00
ArgusAI AI Data Bill of Materials (DBOM) Not checked yet new 43

Adds ArgusAI's AI Data Bill of Materials (DBOM), which automatically links Amazon Bedrock Custom Models, Knowledge Bases (RAGRAGRetrieval-augmented generation: looking up relevant documents and adding them to the prompt before the model answers. It grounds responses in private or current data without retraining.), and Agents to the datasets used during training and inference.

— States what it does, no mechanism or usage steps launch-20260929-1fe46b0a
10 00
Metadata Lake Copilot Not checked yet new 38

Launches Bedrock Metadata Lake Copilot, a natural-language interface for querying the metadata lake to identify sensitive assets, riskiest users, and other data-risk questions.

— Brief description, no example query or navigation launch-20260929-d9dffb84
11 00
Metadata Lake API integration and detection capabilities Not checked yet new 38

Exposes an API-first architecture enabling teams to implement and enforce uniform data-security policies across heterogeneous enterprise infrastructure and to integrate the metadata lake with SIEMsSIEMsSecurity Information and Event Management platforms that collect, normalize, and correlate log data across an infrastructure so cyber tools can detect threats and trigger alerts from a central point., CNAPPs, and DLPs. Also supports detection of sensitive data used in AI initiatives and plain-text secrets, and provides external sharing discovery.

— Lists capabilities thinly, no endpoint or config named launch-20260929-d9dffb84

Bedrock Data launches Agent DLPDLPData Loss Prevention: a category of controls that inspect and block sensitive data from leaving an organisation via email, web, or removable media, giving cyber tools a policy enforcement layer over data in motion and at rest. inside ArgusAI — inline runtime inspection of every AI agent tool call via AWS AgentCore and LiteLLM hooks.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

  • Adds Agent DLP to ArgusAI with native hooks for AWS AgentCore and LiteLLM, sitting inline at the agent gateway to inspect every tool call in both directions without acting as a proxy.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠ Not checked yet
  • Inspects both outbound agent requests to 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. tools and inbound tool responses, enforcing data access and regulatory policies to mask, redact, or block sensitive data exposure in real time. Not checked yet
  • Logs every tool call decision with its target, action, data types involved, and verdict (blocked, modified, redacted, or observed), producing a continuous audit-ready record of agent activity. Not checked yet
  • Supports an observe mode that logs agent actions — such as a marketing agent returning customer emails — without blocking, for low-risk or monitoring-only enforcement scenarios. Not checked yet
  • Extends ArgusAI posture coverage (data store inventory, identity mapping, natural-language access policies) to runtime enforcement, closing the gap between declared policy and actual agent behavior. Not checked yet
  • Adds Agent DLP to ArgusAI with native hooks for AWS AgentCore and LiteLLM, sitting inline at the agent gateway to inspect every tool call in both directions without acting as a proxy.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠ Not checked yet
  • Inspects both outbound agent requests to 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. tools and inbound tool responses, enforcing data access and regulatory policies to mask, redact, or block sensitive data exposure in real time. Not checked yet
  • Logs every tool call decision with its target, action, data types involved, and verdict (blocked, modified, redacted, or observed), producing a continuous audit-ready record of agent activity. Not checked yet
  • Supports an observe mode that logs agent actions — such as a marketing agent returning customer emails — without blocking, for low-risk or monitoring-only enforcement scenarios. Not checked yet

1 more in Everything

Bedrock Data launches ArgusAI to link AI models to their data lineage, assess guardrail gaps, and enforce plain-English policies across AI systems.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

└──▷ HOW TO FIND IT
Investigate which AI models in your Amazon Bedrock environment can access protected health information, using plain-English natural language queries.
📍In the Bedrock Data Platform, navigate to ArgusAI › Natural Language Investigation and enter: 'Which models can access PHI data?' — the system returns results with lineage and access evidence from the Metadata Lake.
Enforce a data residency policy across AI systems without writing control-plane code — useful for EU compliance requirements.
📍In the Bedrock Data Platform, go to Natural Language Policy and enter: 'EU personal data must remain within EU regions.' — Bedrock Data translates the invariant into enforceable controls across connected systems including AI agents and Knowledge Bases.
  • Adds ArgusAI, a new product that automatically generates an AI Data Bill of MaterialsBill of MaterialsA structured list of every component, library, and dependency that makes up a piece of software or hardware, used by cyber tools to track what is inside a product and surface known vulnerabilities in its parts. (DBOM) by linking Amazon Bedrock Custom Models, Knowledge Bases (RAGRAGRetrieval-augmented generation: looking up relevant documents and adding them to the prompt before the model answers. It grounds responses in private or current data without retraining.), and Agents to the datasets used during training and inference.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠ Not checked yet
  • Adds Guardrail Gap Analysis that compares policy intent against model data exposure — drawn from the Metadata Lake — and generates targeted remediations to suppress or mask sensitive data at the source or agent runtime. Not checked yet
  • Adds Natural Language Policy and Investigation, enabling teams to articulate data controls in plain English (e.g., 'Only members of HR may view employment data' or 'EU personal data must remain within EU regions') and have them translated into enforceable controls across connected systems. Not checked yet
  • Extends Natural Language Policy over the DBOM and guardrail context so investigators can query AI governance in plain English — for example, 'Which models can access PHIPHIProtected Health Information: individually identifiable health data regulated under the US HIPAA law, whose strict handling requirements drive cyber tools to audit, mask, or restrict access to it. data?' or 'Which identities interacted with customer 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. through an agent?' — with answers derived from the Metadata Lake including lineage and access evidence. Not checked yet

4 details in Everything

Bedrock Security launches a Metadata Lake with Copilot, AI Agents for DSPMDSPMData security posture management, locating sensitive data and checking how it is protected and who can reach it. It helps teams prioritize risks around their data stores., and a free Snowflake edition for data discovery and classification.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

  • Introduces the Bedrock Metadata Lake, a graph-backed centralized repository that continuously catalogs enterprise data stores across SaaSSaaSSoftware delivered over the internet and billed by subscription, hosted and operated by the vendor so the buyer skips installation and infrastructure management., PaaS, and IaaS, capturing location, sensitivity, entitlements, lineage, usage, risk, and more than fifty additional parameters.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠ Not checked yet
  • Launches Bedrock Metadata Lake Copilot, a natural-language interface for querying the metadata lake to identify sensitive assets, riskiest users, and other data-risk questions. Not checked yet
  • Introduces Bedrock AI Agents for DSPM, which proactively analyze the metadata lake for policy violations and new risks, gather context automatically, suggest next steps, and guide users through incident-response resolution workflows. Not checked yet
  • Launches Bedrock Free for Snowflake, a free tier that delivers automated sensitive-data discovery and classification for Snowflake environments, including access to Metadata Lake Copilot, with upgrade paths to entitlement chain and usage analysis and automated policy capabilities. Not checked yet
  • Exposes an API-first architecture enabling teams to implement and enforce uniform data-security policies across heterogeneous enterprise infrastructure and integrate the metadata lake with SIEMsSIEMsSecurity Information and Event Management platforms that collect, normalize, and correlate log data across an infrastructure so cyber tools can detect threats and trigger alerts from a central point., CNAPPs, and DLPs. Not checked yet
+1 more show less
  • Supports detection of sensitive data used in AI initiatives and plain-text secrets, and provides external sharing discovery, through the metadata lake platform. Not checked yet

6 details in Everything

Bedrock Security announces an 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. providing standardized AI-agent access to its metadata lake, covering 50+ data-context signals.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

  • Adds an MCP (Model Context ProtocolModel Context ProtocolModel 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 as a standardized gateway between AI agents and Bedrock's metadata lake, exposing data sensitivity, risk, lineage, access, and usage signals via the open MCP protocol.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠ Not checked yet
  • Surfaces over 50 distinct metadata signals — including sensitivity classification, access patterns, and lineage — as queryable context for agentic AI workflows and automated governance pipelines. Not checked yet
  • Enables AI agents and external applications to query the metadata lake without custom point-to-point integrations, allowing teams such as legal and compliance to run data investigations without engineering bottlenecks. Not checked yet
  • Builds on the existing Bedrock Metadata Lake Copilot, extending its natural-language metadata querying to agentic and automated workflow contexts. Not checked yet
  • Audits AI model interactions with enterprise data through the MCP Server to support traceability and policy alignment in agentic workflows. Not checked yet
  • Adds an MCP (Model Context ProtocolModel Context ProtocolModel 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 as a standardized gateway between AI agents and Bedrock's metadata lake, exposing data sensitivity, risk, lineage, access, and usage signals via the open MCP protocol.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠ Not checked yet
  • Surfaces over 50 distinct metadata signals — including sensitivity classification, access patterns, and lineage — as queryable context for agentic AI workflows and automated governance pipelines. Not checked yet
  • Enables AI agents and external applications to query the metadata lake without custom point-to-point integrations, allowing teams such as legal and compliance to run data investigations without engineering bottlenecks. Not checked yet
  • Audits AI model interactions with enterprise data through the MCP Server to support traceability and policy alignment in agentic workflows. Not checked yet

1 more in Everything

Bedrock Security extends Agent DLPDLPData Loss Prevention: a category of controls that inspect and block sensitive data from leaving an organisation via email, web, or removable media, giving cyber tools a policy enforcement layer over data in motion and at rest. to NVIDIA OpenShell, adding data-aware policy enforcement at the agent sandbox boundary.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

  • Extends Agent DLP to NVIDIA OpenShell via OpenShell's Supervisor Middleware, evaluating each agent action against enterprise data policies and returning an allow/block decision that OpenShell enforces at the sandbox boundary.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠ Not checked yet
  • Supports an observe-only mode that records policy decisions without enforcing them, letting teams validate coverage before turning on enforcement. Not checked yet
  • At design time, exposes an agent's blast radius — what data its permissions can reach, what combinations that data forms, and where policy would be violated — giving security teams pre-deployment visibility. Not checked yet
  • Applies the same data policies across every tool invocation path an agent may use — 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. tools, command-line invocations, and code the agent writes itself — from a single OpenShell control point. Not checked yet
  • Carries data restrictions across table writes, tool handoffs, and agent-to-agent transfers based on the Metadata Lake knowledge graphknowledge graphA representation of entities and the relationships between them. Software can follow those connections to answer questions or combine information from different sources. rather than payload pattern matching, so restrictions travel with the data as it moves. Not checked yet

5 details in Everything

Good pick?
Good reason?

Where each pick stands · Sep 25–Oct 1

Tools with changes in each pick's category over the last 12 weeks, including the pick itself.

Each pick's release rhythm · Sep 25–Oct 1

Changes per week for the last 12 weeks. The bright bar is this week; a dash marks a week with no daily.

Which categories moved · Sep 25–Oct 1

The thick bar is this week. The thin bar is the average of the four weeks before, counting only weeks with a daily.

  • Identity & Access Management 94usual 90.8
  • Code Security 77usual 104.5
  • SIEM 46usual 42
  • Secure Access (VPN / SASE / ZTNA) 39usual 21.2
  • Endpoint Security (EDR/XDR) 37usual 19.8
  • Model & Agent Protection 33usual 21.8
  • Secrets Detection & Management 33usual 19.8
  • Data Security & DLP 29usual 18.5
  • Internet Services 26usual 14
  • General Utilities & Libraries 25usual 19.8
  • Web Application Security 25usual 28
  • MCP Servers & Tooling 22usual 5.5

Who shipped the most · Sep 25–Oct 1

Changes per tool this week. Volume is not a quality score.

Saved in this browser. Sign in to sync your tools.