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 013, September 25, 2026

THE CYBER TOOLCHAIN NO. 013
Head
THE WEEK'S TOP PICKS
PUBLISHED SEPTEMBER 25, 2026 · FRIDAYS
EDITIONS tail grep head ls diff uniq

The best tooling updates that shipped this week.

The week in numbers · Sep 18–24

Changes per week, last 12 weeks · Sep 18–24

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

81
Jul 3
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

Category moves this month

Checking Plays…

Every feature release from the 1325 tools on our watchlist goes into the daily newsletter. Once a week we read the whole field side by side and choose our top picks, judged on two questions: how deep and complete is the single best capability in the release, and how much it changes what you can actually do.

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 SonarQube announced · Sep 21, 2026 Code Security

Trusted by 7M+ developers, SonarQube is the AI code review and code verification platform that catches quality and security issues before they reach production.

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

SonarQube Hunter Agent adds AI-driven detection of business-logic, broken access control, and authentication/session flaws that static analysis and SCA usually miss. It lands findings as normal SonarQube issues with severity, file, line, triage, scheduling, and an auditable evidence trail, making it immediately usable in existing AppSec workflows.

SonarQube Cloud added first-class Rust analysis, including ClippyClippyA Rust crate, maintained by the Rust project, that adds extra lint checks to catch common mistakes, non-idiomatic code, and suspicious patterns before they reach production.-based rules, Rust metrics, coverage, duplication detection, quality gates, and scanner properties for CargoCargoRust’s package manager and build tool, maintained by the Rust project. Cyber tools use it to fetch dependencies, compile crates, run tests, and produce reproducible builds. and Clippy control.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠ The window also brought Hunter Agent AI verification, native R analysis, and zero-configuration GitLab automatic analysis.

└──▷ WHAT SHIPPED · 4 FEATURES 1 OF 4 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
Rust analysis in SonarQube Cloud new 96

SonarQube Cloud now supports Rust as a first-class language with 122 curated Clippy rules enforced through quality profiles, plus library- and framework-aware Rust rules for tokio async handling, serde serialization, and actix-web misuses.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠ It calculates Cognitive Complexity and Cyclomatic Complexity, imports LCOVLCOVAn open-source front end to GCC's gcov coverage tool that collects line and branch hit counts from instrumented C/C++ code and renders them as HTML reports, giving cyber tools a standard way to measure test coverage. or Cobertura coverage into the issue view, detects duplicate Rust code with CPD, analyzes Rust in the same multi-language project view as C/C++, Java, and JavaScript, and supports quality gate enforcement for Rust with pass/fail thresholds scoped to new code for CI blocking. The scanner adds sonar.rust.cargo.manifestPaths for non-default Cargo.toml locations and sonar.rust.clippy.enabled, which can be set to false to import an external Clippy JSON report instead of running cargo clippy automatically.

Analyze a Rust workspace where Cargo.toml is not at the repo root — point the scanner at the correct manifest path.
$ sonar-scanner -Dsonar.rust.cargo.manifestPaths=services/myservice/Cargo.toml
Use a pre-generated Clippy JSON report (e.g. from a prior CI step) instead of letting SonarQube run cargo clippy itself.
$ sonar-scanner -Dsonar.rust.clippy.enabled=false
— Exact scanner properties and runnable examples launch-20260922-39805485
02 01
Hunter Agent AI verification new 74

Hunter Agent adds an AI-driven verification layer that detects business-logic, broken access control, and authentication/session vulnerabilities through structured multi-stage playbooks. Findings appear as standard SonarQube issues with severity, file, and line in the existing triage, assignment, and tracking workflow, are summarized on a dedicated Hunter Agent page, can run manually or daily against long-lived branches, and produce an auditable evidence trail mapped to DORA, 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., PCI DSSPCI DSSThe Payment Card Industry Data Security Standard, security requirements for protecting payment account data. It covers organizations that store, process or transmit that data and relevant service providers. 4.0.1, 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., GDPRGDPREU regulation enacted in 2018 that governs how organizations collect, store, and process personal data of EU residents, giving cyber tools legal obligations around data handling and breach notification., EU CRAEU CRAEU Cyber Resilience Act, a European Union regulation requiring manufacturers and publishers of products with digital elements to meet mandatory cybersecurity standards before selling into the EU market., NYDFS, and FDA 21 CFR Part 11 requirements.

— Strong behavior detail, limited operation path launch-20260921-74341ed0
03 00
GitLab automatic analysis new 67

SonarQube Cloud extends zero-configuration automatic analysis, previously available for GitHub and Azure DevOps, to GitLab. It scans eligible GitLab repositories without .gitlab-ci.yml edits, covers 20+ languages including C#, Java, and Python, and starts analysis after users link GitLab, import repositories, and select eligible repositories.

Onboard a GitLab repository to SonarQube Cloud for instant, pipeline-free scanning on every merge request.
📍1. Sign in to SonarQube Cloud and link your GitLab account (ensure your access token has repository and metadata read permissions). 2. Go to Import repositories and select the GitLab repositories you want to analyze. 3. For eligible repositories, automatic analysis starts immediately — no further configuration needed. 4. Results appear on the default branch and active merge requests shortly after import.
— Clear onboarding steps, limited internals launch-20260915-5d8d3f1f
04 00
Native R language analysis new 65

SonarQube adds native R support on Cloud and Server, analyzing R source files and embedded R code inside R Markdown documents with 82 rules covering key sections of the lintr rule set plus Sonar-developed rules. It imports lintr problem reports, adds Cobertura test coverage import, extends secrets detection, code metrics, syntax highlighting, and copy-paste duplication analysis to R, and applies existing quality profiles, quality gates, PR decoration, and branch analysis to R codebases.

— Names formats and scope, no command launch-20260915-f7587a71

SonarQube Hunter Agent adds AI-driven detection of business logic, access control, and authentication flaws as standard issues.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

  • Adds Hunter Agent, a new AI-driven verification layer that detects business-logic, broken access control, and authentication/session vulnerabilities through structured multi-stage playbooks — classes of flaw that static analysis and 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. miss.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠
  • Surfaces Hunter Agent findings as standard SonarQube issues (with severity, file, and line) in the existing triage, assignment, and tracking workflow, with results also summarized on a dedicated Hunter Agent page.
  • Supports running Hunter Agent manually or on a daily schedule against long-lived branches to suit different team cadences.
  • Produces a reproducible, auditable evidence trail from detection through triage and remediation, mapped to DORA, 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., PCI DSSPCI DSSThe Payment Card Industry Data Security Standard, security requirements for protecting payment account data. It covers organizations that store, process or transmit that data and relevant service providers. 4.0.1, 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., GDPRGDPREU regulation enacted in 2018 that governs how organizations collect, store, and process personal data of EU residents, giving cyber tools legal obligations around data handling and breach notification., EU CRAEU CRAEU Cyber Resilience Act, a European Union regulation requiring manufacturers and publishers of products with digital elements to meet mandatory cybersecurity standards before selling into the EU market., NYDFS, and FDA 21 CFR Part 11 requirements.

4 details in Everything

SonarQube Cloud adds first-class Rust support: 122 Clippy rules, complexity metrics, coverage import, and quality gates.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

└──▷ TRY IT
Analyze a Rust workspace where Cargo.toml is not at the repo root — point the scanner at the correct manifest path.
$ sonar-scanner -Dsonar.rust.cargo.manifestPaths=services/myservice/Cargo.toml
Use a pre-generated Clippy JSON report (e.g. from a prior CI step) instead of letting SonarQube run cargo clippy itself.
$ sonar-scanner -Dsonar.rust.clippy.enabled=false
  • Adds sonar.rust.cargo.manifestPaths config property to point the Rust analyzer at non-default Cargo.toml locations.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠
  • Adds sonar.rust.clippy.enabled config property (set to false to import an external Clippy JSON report instead of running cargo clippy automatically).
  • Supports Rust as a first-class language in SonarQube Cloud, with 122 curated Clippy rules enforced through quality profiles across all projects.
  • Adds library- and framework-aware Rust rules covering tokio async handling, serde serialization, and actix-web misuses — beyond what Clippy detects.
  • Adds Cognitive Complexity and Cyclomatic Complexity metric calculation for Rust code.
+4 more show less
  • Imports Rust coverage data in LCOVLCOVAn open-source front end to GCC's gcov coverage tool that collects line and branch hit counts from instrumented C/C++ code and renders them as HTML reports, giving cyber tools a standard way to measure test coverage. or Cobertura format and correlates it with analysis results in the issue view.
  • Adds duplicate code detection (CPD) for Rust, catching copy-pasted logic across the codebase.
  • Analyzes Rust inside the same multi-language project view as C/C++, Java, and JavaScript, with no separate dashboard required.
  • Supports quality gate enforcement for Rust, including pass/fail thresholds scoped to new code for CI blocking.

9 details in Everything

SonarQube adds native R language support with 82 rules, secrets detection, lintr import, and CoberturaCoberturaAn open-source XML format for reporting code-coverage results, originally from a Java tool of the same name; cyber tools adopt it because most CI systems can parse and display it natively. coverage on Cloud and Server.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

  • Analyzes R source files and embedded R code inside R Markdown documents with 82 rules covering key sections of the lintr rule set plus additional Sonar-developed rules.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠
  • Imports lintr problem reports directly into SonarQube, surfacing existing findings in the centralized view alongside other languages.
  • Adds Cobertura test coverage import for R projects.
  • Extends secrets detection, code metrics, syntax highlighting, and copy-paste duplication analysis to R.
  • Applies existing quality profiles, quality gates, PR decoration, and branch analysis to R codebases.
  • Analyzes R source files and embedded R code inside R Markdown documents with 82 rules covering key sections of the lintr rule set plus additional Sonar-developed rules.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠
  • Imports lintr problem reports directly into SonarQube, surfacing existing findings in the centralized view alongside other languages.
  • Adds Cobertura test coverage import for R projects.
  • Extends secrets detection, code metrics, syntax highlighting, and copy-paste duplication analysis to R.

1 more in Everything

SonarQube Cloud adds zero-configuration automatic analysis for GitLab, scanning every merge request and main-branch commit without .gitlab-ci.yml edits.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

└──▷ HOW TO FIND IT
Onboard a GitLab repository to SonarQube Cloud for instant, pipeline-free scanning on every merge request.
📍1. Sign in to SonarQube Cloud and link your GitLab account (ensure your access token has repository and metadata read permissions). 2. Go to Import repositories and select the GitLab repositories you want to analyze. 3. For eligible repositories, automatic analysis starts immediately — no further configuration needed. 4. Results appear on the default branch and active merge requests shortly after import.
  • Extends zero-configuration automatic analysis — previously exclusive to GitHub and Azure DevOps — to GitLab, covering 20+ languages including C#, Java, and Python.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

1 detail in Everything

Good pick?
Good reason?
02 Palo Alto Networks changelog · Sep 22, 2026 Firewalls & Network Defense

Palo Alto Networks is a cybersecurity platform that provides advanced firewalls, cloud security, and AI-driven threat detection to protect networks and enterprises.

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

Unit 42 Continuous Frontier AI Defense adds always-on agentic offensive testing that continuously finds, validates, and remediates exposures across web apps, APIs, cloud infrastructure, source repositories, and network assets. The Advanced Adversary Simulation validates end-to-end attack paths, closing a practical gap between asset scanning and proven exploitability.

Palo Alto Networks launched Unit 42 Continuous Frontier AI Defense, an always-on agentic offensive cyber service for continuously finding, validating, and remediating enterprise exposures.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠ The service includes continuous estate testing, multi-model AI routing, and adversary simulation across web apps, APIs, cloud infrastructure, source code repositories, and network assets.

└──▷ WHAT SHIPPED · 4 FEATURES 4 OF 4 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
Multi-Model AI Harness new 62

The proprietary Multi-Model AI Harness routes offensive security tasks to the model best suited for each job.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠ Palo Alto Networks says this addresses findings that no single model catches more than 40% of vulnerabilities and that leading models share less than 10% overlap in identified exposures.

— Mechanism and metrics, no operator controls 2026-09-22-d8eee514
02 02
Continuous Frontier AI Defense new 60

Unit 42 Continuous Frontier AI Defense is a new always-on agentic offensive security service that continuously finds, validates, and remediates enterprise exposures. It uses proprietary multi-model AI harnesses with Anthropic Claude Mythos 5 and OpenAI GPT-5.6-Cyber.

— Names service and models, but no usage surface 2026-09-22-d8eee514
thinner coverage below
03 03
Continuous Testing Engine new 59

The Continuous Testing Engine performs a full-estate baseline scan followed by always-on testing as an organization's environment changes. Coverage includes web apps, APIs, cloud infrastructure, source code repositories, and network assets.

— Clear scope, no access path 2026-09-22-d8eee514
04 04
Advanced Adversary Simulation new 53

Advanced Adversary Simulation validates end-to-end attack pathsattack pathsA 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. to prove real-world exploitability. It works across first- and third-party web apps, APIs, cloud infrastructure, source code repositories, and network assets.

— Explains validation scope, not how to run 2026-09-22-d8eee514

Palo Alto Networks launches Unit 42 Continuous Frontier AI Defense, an always-on agentic offensive security service powered by Anthropic Claude Mythos 5 and OpenAI GPT-5.6-Cyber.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

  • Introduces Unit 42 Continuous Frontier AI Defense, a new always-on agentic offensive security service that continuously finds, validates, and remediates enterprise exposures using proprietary multi-model AI harnesses with Anthropic Claude Mythos 5 and OpenAI GPT-5.6-Cyber.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠
  • Adds a Continuous Testing Engine that performs a full-estate baseline scan followed by always-on testing as an organization's environment changes, covering web apps, APIs, cloud infrastructure, source code repositories, and network assets.
  • Introduces a proprietary Multi-Model AI Harness that routes offensive security tasks to the model best suited for each job, addressing the finding that no single model catches more than 40% of vulnerabilities and that leading models share less than 10% overlap in identified exposures.
  • Delivers Advanced Adversary Simulation that proves real-world exploitability by validating end-to-end attack pathsattack pathsA 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. across first- and third-party web apps, APIs, cloud infrastructure, source code repositories, and network assets.

4 details in Everything

Good pick?
Good reason?
03 Semgrep announced · Sep 24, 2026 Code Security

Semgrep is a static analysis tool that scans code for bugs, security vulnerabilities, and policy violations using customizable rules.

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

Semgrep Supply Chain adds zero-day malware detection that rescans customer repositories within minutes of a supply-chain incident and reports affected repos, projects, findings, and scan progress. The new org-wide response policies and Malware Firewall extend that from detection into automated alerting and pre-install blocking.

Semgrep Supply Chain added zero-dayzero-dayA vulnerability for which defenders have had no prior opportunity to apply a fix, often because it was previously unknown. A zero-day exploit takes advantage of that gap. malware detection and response automation, including org-wide policies, incident rescans, early blast-radius notifications, and a private-beta Malware Firewall.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠ Semgrep also added CSV export for Agentic Workflows issues from the Issues page and public API.

└──▷ WHAT SHIPPED · 5 FEATURES 4 OF 5 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
Malware Firewall private beta new 75

Semgrep added a Malware Firewall in private beta, embedded in Semgrep Guardian via Agent admin.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠ It intercepts dependency requests at the network level between the package manager and public registries and blocks malicious packages before they reach developer machines, with no per-device install.

— Mechanism and placement are described launch-20260924-77b2b3de
02 02
Supply Chain malware response automation new 65

Semgrep Supply Chain now supports org-wide policies that trigger automated responses when a malware finding is generated, including opening a Jira ticket, sending a Slack alert, or calling a webhook with no human in the loophuman in the loopHuman in the loop, involving a person in an otherwise automated workflow. The person can review, approve, correct or supply information before work continues..

— Names response targets, but no setup path launch-20260924-77b2b3de
03 03
Supply Chain incident notifications new 64

Customers can opt in to early notifications that proactively report their blast radius via Slack or webhooks when a new supply chain incident is detected. Notifications cover affected repos, projects, findings, and scan progress.

— Clear channels and contents, no configuration details launch-20260924-77b2b3de
04 04
Zero-day incident detection new 63

Semgrep added zero-day incident detection that rescans all customer repositories with AI-assisted, expert-verified advisory rules within minutes of a supply chain incident. Findings are posted to the Semgrep dashboard, organized by advisories and rolled up into incidents.

— Explains behavior, but no user control launch-20260924-77b2b3de
thinner coverage below
05 00
Agentic Workflows CSV export new 49

Semgrep now supports exporting Agentic Workflows issues as CSV from the Issues page and through the public API.

— Names format and entry points only snapshot-20260923

Semgrep Supply Chain adds zero-dayzero-dayA vulnerability for which defenders have had no prior opportunity to apply a fix, often because it was previously unknown. A zero-day exploit takes advantage of that gap. malware detection with automated org-wide response policies and a malware firewall (private beta).⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

  • Adds org-wide Supply Chain policies that trigger automated responses — open a Jira ticket, send a Slack alert, or call a webhook — the moment a malware finding is generated, with no human in the loophuman in the loopHuman in the loop, involving a person in an otherwise automated workflow. The person can review, approve, correct or supply information before work continues..⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠
  • Adds zero-day incident detection that rescans all customer repositories with AI-assisted, expert-verified advisory rules within minutes of a supply chain incident and posts findings to the Semgrep dashboard organized by advisories rolled up into incidents.
  • Adds early-notification opt-in that proactively informs customers via Slack or webhooks of their blast radius — affected repos, projects, findings, and scan progress — when a new supply chain incident is detected.
  • Adds a Malware Firewall (private beta), embedded in Semgrep Guardian via Agent admin, that intercepts dependency requests at the network level between the package manager and public registries and blocks malicious packages before they reach developer machines, with no per-device install.
  • Adds org-wide Supply Chain policies that trigger automated responses — open a Jira ticket, send a Slack alert, or call a webhook — the moment a malware finding is generated, with no human in the loophuman in the loopHuman in the loop, involving a person in an otherwise automated workflow. The person can review, approve, correct or supply information before work continues..⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠
  • Adds early-notification opt-in that proactively informs customers via Slack or webhooks of their blast radius — affected repos, projects, findings, and scan progress — when a new supply chain incident is detected.
  • Adds a Malware Firewall (private beta), embedded in Semgrep Guardian via Agent admin, that intercepts dependency requests at the network level between the package manager and public registries and blocks malicious packages before they reach developer machines, with no per-device install.

1 more in Everything

Semgrep adds CSV export for Agentic Workflows issues via the Issues page and public API.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

Good pick?
Good reason?
04 Docker Sandboxes announced · Sep 24, 2026 Model & Agent Protection
// WHY IT MADE THE LIST Depth 5/5 Impact 5/5

Docker Sandboxes adds Cloud Sandboxes and an OCI-native Kit format that lets teams run coding agents in managed microVMs with declared network and credential authority. Kits, proxy-managed secrets, network policies, and authority-diff gating make agent execution governable rather than just isolated.

Docker Sandboxes launched managed Cloud Sandboxes and the v3 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. Kit format for packaging sandbox authority.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠ The window also added model-provider configuration, 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. transport controls, network-policy hardening, richer CLI/JSON surfaces, and several breaking CLI behavior changes.

└──▷ WHAT SHIPPED · 26 FEATURES 3 OF 26 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
OCI v3 Kit specification new 96

Docker Sandbox Kit Specification v3 defines a portable OCI image format, open source under Apache 2.0 at docker/sandbox-kit-spec, for declaring what authority a sandbox agent may exercise.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠ Kit declarations live in the OCI manifest annotation vnd.docker.sandbox.kit.descriptor, build with docker buildx build, pull with docker pull, and are pinnable by digest; v3 kits separate workload Kits, which supply the root filesystem, from mixin Kits, which overlay a CLI, network rule, or credential binding. The spec defines versioned capability types including com.docker.sandbox/network-policy@2 and com.docker.sandbox/credential@1, plus conformance suites for artifacts and runtimes.

Run a sandboxed agent with the GitHub CLI mixin to give it scoped GitHub access defined by the Kit's network-policy and credential declarations.
$ cd examples
sbx run ./hello --kit ./gh
Author a mixin Kit that allows most GitHub API methods but explicitly denies DELETE on /repos/**, keeping a token that can open PRs from deleting repositories.
yaml
capabilities:
  - type: com.docker.sandbox/network-policy@2
    config:
      runtime:
        allow:
          - github.com
          - hosts: [api.github.com]
            methods: [GET, HEAD, POST, PATCH, PUT, DELETE]
        deny:
          - hosts: [api.github.com]
            methods: [DELETE]
            paths: [/repos/**]
  - type: com.docker.sandbox/credential@1
    optional: true
    config:
      service: github
      phase: runtime
      apiKey:
        name: GH_TOKEN
      proxyManaged: true
      inject:
        - {domain: api.github.com, header: Authorization, format: 'Bearer %s'}
— Exact spec surfaces and build commands v0.45.0launch-20260924-0c0143b0
02 02
Cloud Sandboxes compute new 94

Cloud Sandboxes run sandboxes on Docker-managed cloud compute with sbx --cloud run <kit> and support moving local work with sbx move --to cloud. Size options range from Micro (1 vCPU / 2 GiB, $0.07/hr) to XL (16 vCPU / 32 GiB, $1.12/hr), sessions can run up to 24 hours, and cloud support requires sbx version 0.45.1 or later.

Launch a Claude Code agent in the cloud before leaving for the day so it keeps running overnight without your laptop.
$ sbx --cloud run claude
Hand off a long-running refactor you started locally to cloud compute so you can close your laptop without interrupting the agent.
$ sbx move my-project --to cloud
— Commands, limits, pricing, version requirement launch-20260924-6c1aef17
03 03
Kit authority enforcement new 93

V3 Kit capabilities can scope network access by HTTP method and path, declare install-phase network access and credential use, and specify where an agent reads shared skills. Network-policy capabilities support explicit allow and deny rules scoped to hosts, methods, and paths, with deny taking precedence, while credential capabilities support proxyManaged: true, injecting real tokens into requests to named domains and exposing only a sentinel inside the sandbox. Authority-diff gating reduces every descriptor to a normalized grant set so a runtime may auto-apply updates that stay within the recorded set, but any widening, including removing a deny rule, requires explicit approval.

Author a mixin Kit that allows most GitHub API methods but explicitly denies DELETE on /repos/**, keeping a token that can open PRs from deleting repositories.
yaml
capabilities:
  - type: com.docker.sandbox/network-policy@2
    config:
      runtime:
        allow:
          - github.com
          - hosts: [api.github.com]
            methods: [GET, HEAD, POST, PATCH, PUT, DELETE]
        deny:
          - hosts: [api.github.com]
            methods: [DELETE]
            paths: [/repos/**]
  - type: com.docker.sandbox/credential@1
    optional: true
    config:
      service: github
      phase: runtime
      apiKey:
        name: GH_TOKEN
      proxyManaged: true
      inject:
        - {domain: api.github.com, header: Authorization, format: 'Bearer %s'}
— Concrete capability fields and enforcement behavior v0.45.0launch-20260924-0c0143b0
04 00
Network policy protocol selection new 88

Local network allow rules gained --protocol, allowing 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. to be selected explicitly alongside the default TCPTCPA network protocol that delivers a reliable, ordered stream of bytes between two endpoints. It detects lost data and retransmits it. coverage. Experimental outbound UDP now follows sandbox network policy.

Allow outbound UDP (e.g. for a DNS-over-UDP workflow) in a sandbox network policy instead of the TCP-only default.
$ sbx policy allow network --protocol udp 8.8.8.8:53
— Exact flag and behavior v0.45.0
05 00
Network policy bypass fixes improved 87

Reverse-DNS lookups are now restricted to destination IPs already authorized by policy, closing a previous policy bypass. DNS resolution is blocked when no network rule permits it, and sbx policy allow network, sbx policy deny network, --allow-network, and --deny-network now reject malformed patterns before saving them.

— Commands and before-after policy behavior v0.45.0
06 00
Non-interactive destructive confirmations breaking 86

Destructive commands gained --force for non-interactive confirmation, and sbx kit builder history rm also gained --yes and -y. Commands that remove resources now prompt for confirmation and return a non-zero exit code when the prompt is declined.

— Flags and changed exit behavior v0.45.0
07 00
MCP catalog removal breaking 86

sbx mcp catalog has been removed. Register a remote MCP server with sbx mcp add and then authorize it with sbx mcp auth instead.

— Removed command and replacement commands v0.45.0
08 00
Custom model providers new 85

sbx run --model now supports any OpenAI- or Anthropic-compatible endpoint configured through the new model.providers setting. For sbx run opencode --model, the model's supported thinking levels are exposed as OpenCode variants selectable with Ctrl+T.

Route agent runs through a custom Anthropic-compatible endpoint instead of the built-in provider.
$ sbx run --model my-anthropic-proxy/claude-3-5-sonnet 'summarize the repo'
— Flag, setting, compatible endpoints v0.45.0
09 00
Remote MCP HTTP/2 control new 84

sbx mcp add --disable-http2 disables HTTP/2HTTP/2The second major version of the HTTP protocol, standardized by the IETF, adding multiplexed streams, header compression, and binary framing over a single TCP connection to cut latency and reduce overhead. for a remote MCP transport. The flag is intended for servers whose HTTP/2 handling stalls long-lived streams.

Work around a remote MCP server that stalls long-lived HTTP/2 streams by forcing HTTP/1.1.
$ sbx mcp add --disable-http2 my-mcp-server https://mcp.example.com/sse
— Exact command flag and use case v0.45.0
10 00
Missing resource errors breaking 82

sbx secret rm now returns an error on stderr when the requested secret does not exist, where it was previously silent. sbx mcp rm now returns an error when the requested MCP server is not registered, where it was previously silent.

— Exact commands and changed failure mode v0.45.0
11 00
Kit validation hardening improved 80

sbx kit validate now rejects malformed API-key declarations, including invalid names, missing injection domains, invalid format placeholders, and Basic-auth usernames containing a colon. It also warns for no-op or out-of-allowlist declarations.

— Validator command and rejection cases v0.45.0
12 00
Detached exec failure breaking 79

sbx exec -d and --detach now fail immediately for unsupported detached execution instead of silently running in the foreground.

— Exact flags and before-after behavior v0.45.0
13 00
Kit sets and composition new 77

Kit sets let authors combine a workload and mixins, pin component versions, and publish the result as a single OCI reference with its own capabilities, lifecycle hooks, instructions, and arguments. A kind: set descriptor names a collection of Kits, runs coherence checks at build time, and merges them into one ordinary Kit so incoherence is caught at the author's build rather than at launch.

— Names descriptor field and behavior v0.45.0launch-20260924-0c0143b0
14 00
Sandbox JSON metadata improved 72

sbx ls --json and sbx inspect --json now include created_at, CPUCPUThe main general-purpose processor in a computer. It runs application instructions and coordinates work with memory and other hardware. limits, and memory limits in their output.

— Exact commands and JSON field v0.45.0
15 00
Cloud Kit catalog new 71

Cloud Sandboxes introduced pre-configured, pre-built Kits for leading coding agents, launchable with sbx --cloud run <kit>. Named Kits include Claude Code, Codex, Copilot, Antigravity, Open Code, and Hermes.

Launch a Claude Code agent in the cloud before leaving for the day so it keeps running overnight without your laptop.
$ sbx --cloud run claude
— Command and supported kit names launch-20260924-6c1aef17
16 00
Policy listing creation filter new 67

sbx policy ls gained the --created-via filter. The command also now shows how each rule was created.

— Exact command and filter v0.45.0
17 00
Name-only list output new 65

MCP server, secret, skill, template, volume, and policy-profile list commands gained --quiet and -q for name-only output.

— Exact flags, broad command family v0.45.0
18 00
Signed macOS app bundle improved 60

macOS distributions now ship as a single signed Sbx.app bundle. Homebrew and tarball PATH installs work through a symlink into the bundle.

— Package name and install behavior v0.45.0
thinner coverage below
19 00
Cloud secrets proxy new 59

Cloud Sandboxes added a secrets proxy that injects stored keys or tokens per request so agents do not hold the actual secret value. Separately, local and cloud sandboxes maintain separate secrets, templates, and network policies, so existing local workflows must be reconfigured before moving to cloud.

— Mechanism plus migration caveat launch-20260924-6c1aef17
20 00
Centralized MCP gateway new 58

A centralized MCP gateway lets agents running in cloud, locally, or in other clients such as ChatGPT desktop share one set of connected MCP servers. The launch names Jira, Linear, Grafana, incident.io, and any streamable HTTP endpoint as supported server targets.

— Named integrations, no setup surface launch-20260924-6c1aef17
21 00
OAuth credential coordination improved 56

Multiple OAuth-backed agents can now be composed in the same sandbox with credentials scoped to the individual kits that request them. OAuth refreshes are coordinated across sandboxes that share credentials, preventing simultaneous refreshes from forcing another sign-in.

— Behavior described, no user surface v0.45.0
22 00
V3 kit compatibility break breaking 56

V3 workloads and mixins must be used together. They cannot be combined with v1 or v2 kits.

— Compatibility constraint only v0.45.0
23 00
OpenCode GitHub Copilot setup improved 52

The OpenCode kit now configures GitHub Copilot from the account's stored GitHub credential, enabling Copilot models without a separate device login.

— Integration named, no command v0.45.0
24 00
Cloud network policies new 49

Cloud Sandboxes added network policies that define which endpoints agents can reach. Policies are configured once and applied across all cloud sandboxes, but local and cloud sandboxes maintain separate network policies and require reconfiguration before moving workflows to cloud.

— Scope described, no config syntax launch-20260924-6c1aef17
25 00
Web console for Cloud Sandboxes new 45

A web console can spin up and manage Cloud Sandboxes without using the CLI. The console includes Kit selection and a Run button.

— UI starting point only launch-20260924-6c1aef17
26 00
Global service secret updates improved 45

Adding, updating, or removing global service secrets now updates existing local sandboxes without a restart.

— Clear behavior, no command named v0.45.0
└──▷ BREAKING ON UPGRADE
  • !
    Local and cloud sandboxes maintain separate secrets, templates, and network policies — existing local workflows must be reconfigured before moving to cloud.
  • !
    sbx mcp catalog has been removed; register a remote MCP server with sbx mcp add and then authorize it with sbx mcp auth instead.
  • !
    sbx secret rm now returns an error on stderr when the requested secret does not exist (previously silent).
  • !
    sbx mcp rm now returns an error when the requested MCP server is not registered (previously silent).
  • !
    V3 workloads and mixins must be used together; they cannot be combined with v1 or v2 kits.
  • !
    Commands that remove resources now prompt for confirmation and return a non-zero exit code when the prompt is declined; use --force or --yes/-y in non-interactive workflows.
  • !
    sbx exec -d / --detach now fails immediately for unsupported detached execution instead of silently running in the foreground.

Docker Sandboxes adds Cloud Sandboxes: managed microVM compute in the cloud with sbx --cloud run and sbx move --to cloud.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

└──▷ TRY IT
Launch a Claude Code agent in the cloud before leaving for the day so it keeps running overnight without your laptop.
$ sbx --cloud run claude
Hand off a long-running refactor you started locally to cloud compute so you can close your laptop without interrupting the agent.
$ sbx move my-project --to cloud
  • Adds sbx --cloud run <kit> to launch a sandbox on Docker-managed cloud compute, with size options from Micro (1 vCPU / 2 GiB, $0.07/hr) to XL (16 vCPU / 32 GiB, $1.12/hr) and sessions up to 24 hours.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠
  • Introduces Kits: pre-configured, pre-built sandbox images for leading coding agents (Claude Code, Codex, Copilot, Antigravity, Open Code, Hermes), launchable with a single sbx --cloud run <kit> command.
  • Adds a centralized 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. gateway so agents running in cloud, locally, or in other clients (e.g. ChatGPT desktop) can share a single set of connected MCP servers (Jira, Linear, Grafana, incident.io, or any streamable HTTP endpoint).
  • Adds a Cloud Sandboxes secrets proxy that injects stored keys or tokens per-request so agents never hold the actual secret value.
  • Adds network policies that define which endpoints agents can reach, configurable once and applied across all cloud sandboxes.
+2 more show less
  • Adds a web console for spinning up and managing Cloud Sandboxes without the CLI, including Kit selection and a Run button.
  • Requires sbx version 0.45.1 or later for cloud sandbox support.

7 details in Everything

└──▷ BREAKING ON UPGRADE
  • !Local and cloud sandboxes maintain separate secrets, templates, and network policies — existing local workflows must be reconfigured before moving to cloud.

Docker Sandbox Kit Specification v3 introduces 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.-native Kit format for declaring agent authority as versioned, reviewable config⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

└──▷ TRY IT
Run a sandboxed agent with the GitHub CLI mixin to give it scoped GitHub access defined by the Kit's network-policy and credential declarations.
$ cd examples
sbx run ./hello --kit ./gh
Author a mixin Kit that allows most GitHub API methods but explicitly denies DELETE on /repos/**, keeping a token that can open PRs from deleting repositories.
yaml
capabilities:
  - type: com.docker.sandbox/network-policy@2
    config:
      runtime:
        allow:
          - github.com
          - hosts: [api.github.com]
            methods: [GET, HEAD, POST, PATCH, PUT, DELETE]
        deny:
          - hosts: [api.github.com]
            methods: [DELETE]
            paths: [/repos/**]
  - type: com.docker.sandbox/credential@1
    optional: true
    config:
      service: github
      phase: runtime
      apiKey:
        name: GH_TOKEN
      proxyManaged: true
      inject:
        - {domain: api.github.com, header: Authorization, format: 'Bearer %s'}
  • Introduces the Docker Sandbox Kit Specification v3, open source under Apache 2.0 at docker/sandbox-kit-spec, defining a portable OCI image format for declaring exactly what authority a sandbox agent may exercise.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠
  • Stores all Kit declarations in a single OCI manifest annotation (vnd.docker.sandbox.kit.descriptor), so Kits build with docker buildx build, pull with docker pull, and are pinnable by digest.
  • Supports two Kit kinds: a workload Kit (supplies the root filesystem) and a mixin Kit (an overlay adding a CLI, network rule, or credential binding), composed at launch with one workload and any number of mixins.
  • Defines versioned capability types — including com.docker.sandbox/network-policy@2 and com.docker.sandbox/credential@1 — each with a normative page describing what a conforming runtime must enforce.
  • Network-policy capabilities support explicit allow and deny rules scoped to hosts, methods, and paths, with deny taking precedence, so a mixin can permit most of an API while blocking specific destructive operations (e.g. DELETE /repos/**).
+4 more show less
  • Credential capabilities support proxyManaged: true mode, where the runtime injects the real token into requests to named domains and exposes only a sentinel inside the sandbox.
  • A kind: set descriptor names a collection of Kits and runs coherence checks at build time, merging them into one ordinary Kit so incoherence is caught at the author's build, not at a user's launch.
  • Adds authority-diff gating: every descriptor reduces to a normalized grant set; a runtime may auto-apply updates that stay within the recorded set, but any widening — including removing a deny rule — stops and requires explicit approval.
  • Ships two conformance suites: one that judges whether an artifact is a conforming Kit, and one that verifies whether a runtime implements the specification's normative requirements.

9 details in Everything

Docker Sandboxes v0.45.0 ships v3 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. kits with workload/mixin composition, new model provider settings, and a raft of new CLI flags.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

└──▷ GET THIS VERSION
$ git clone --branch v0.45.0 https://github.com/docker/sbx-releases.git
# already have the repo? check out this version:
$ git checkout v0.45.0
└──▷ TRY IT
Route agent runs through a custom Anthropic-compatible endpoint instead of the built-in provider.
$ sbx run --model my-anthropic-proxy/claude-3-5-sonnet 'summarize the repo'
Allow outbound 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. (e.g. for a DNS-over-UDP workflow) in a sandbox network policy instead of the TCPTCPA network protocol that delivers a reliable, ordered stream of bytes between two endpoints. It detects lost data and retransmits it.-only default.
$ sbx policy allow network --protocol udp 8.8.8.8:53
Work around a remote MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface. server that stalls long-lived HTTP/2HTTP/2The second major version of the HTTP protocol, standardized by the IETF, adding multiplexed streams, header compression, and binary framing over a single TCP connection to cut latency and reduce overhead. streams by forcing HTTP/1.1.
$ sbx mcp add --disable-http2 my-mcp-server https://mcp.example.com/sse
  • Adds sbx run --model support for any OpenAI- or Anthropic-compatible endpoint configured via the new model.providers setting.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠
  • Adds sbx mcp add --disable-http2 to disable HTTP/2 for a remote MCP transport, working around servers whose HTTP/2 handling stalls long-lived streams.
  • Adds --quiet (-q) flag to MCP server, secret, skill, template, volume, and policy-profile list commands for name-only output.
  • Adds --protocol flag to local network allow rules so UDP can be selected explicitly alongside the default TCP coverage.
  • Adds --created-via filter to sbx policy ls, which also now shows how each rule was created.
+16 more show less
  • Adds --force (and --yes/-y for sbx kit builder history rm) flags for non-interactive confirmation of destructive commands.
  • Adds created_at, CPUCPUThe main general-purpose processor in a computer. It runs application instructions and coordinates work with memory and other hardware. limits, and memory limits to sbx ls --json and sbx inspect --json output.
  • Adds sbx kit validate rejection of malformed API-key declarations, including invalid names, missing injection domains, invalid format placeholders, and Basic-auth usernames containing a colon, plus warnings for no-op or out-of-allowlist declarations.
  • Introduces v3 kits: OCI-based packages with separate workload and mixin roles that combine base environments, tools, configuration, credentials, network access, and agent instructions.
  • Introduces kit sets, letting authors combine a workload and mixins, pin component versions, and publish the result as a single OCI reference with its own capabilities, lifecycle hooks, instructions, and arguments.
  • V3 kits can scope network access by HTTP method and path, declare install-phase network access and credential use, and specify where an agent reads shared skills.
  • Multiple OAuth-backed agents can now be composed in the same sandbox with credentials scoped to the individual kits that request them.
  • Experimental outbound UDP now follows sandbox network policy.
  • Adding, updating, or removing global service secrets now updates existing local sandboxes without a restart.
  • OAuth refreshes are coordinated across sandboxes that share credentials, preventing simultaneous refreshes from forcing another sign-in.
  • Reverse-DNS lookups are now restricted to destination IPs already authorized by policy, closing a previous policy bypass.
  • DNS resolution is blocked when no network rule permits it.
  • sbx policy allow network, sbx policy deny network, --allow-network, and --deny-network now reject malformed patterns before saving them.
  • sbx run opencode --model now exposes the model's supported thinking levels as OpenCode variants, selectable with Ctrl+T.
  • The OpenCode kit now configures GitHub Copilot from the account's stored GitHub credential, enabling Copilot models without a separate device login.
  • macOS distributions now ship as a single signed Sbx.app bundle, with Homebrew and tarball PATH installs working via a symlink into the bundle.
  • Adds sbx run --model support for any OpenAI- or Anthropic-compatible endpoint configured via the new model.providers setting.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠
  • Introduces v3 kits: OCI-based packages with separate workload and mixin roles that combine base environments, tools, configuration, credentials, network access, and agent instructions.
  • Introduces kit sets, letting authors combine a workload and mixins, pin component versions, and publish the result as a single OCI reference with its own capabilities, lifecycle hooks, instructions, and arguments.
  • V3 kits can scope network access by HTTP method and path, declare install-phase network access and credential use, and specify where an agent reads shared skills.
  • Multiple OAuth-backed agents can now be composed in the same sandbox with credentials scoped to the individual kits that request them.
+2 more show less
  • Experimental outbound UDP now follows sandbox network policy.
  • OAuth refreshes are coordinated across sandboxes that share credentials, preventing simultaneous refreshes from forcing another sign-in.

14 more in Everything

└──▷ BREAKING ON UPGRADE
  • !sbx mcp catalog has been removed; register a remote MCP server with sbx mcp add and then authorize it with sbx mcp auth instead.
  • !sbx secret rm now returns an error on stderr when the requested secret does not exist (previously silent).
  • !sbx mcp rm now returns an error when the requested MCP server is not registered (previously silent).
  • !V3 workloads and mixins must be used together; they cannot be combined with v1 or v2 kits.
  • !Commands that remove resources now prompt for confirmation and return a non-zero exit code when the prompt is declined; use --force or --yes/-y in non-interactive workflows.
  • !sbx exec -d / --detach now fails immediately for unsupported detached execution instead of silently running in the foreground.
Good pick?
Good reason?

Cycode’s Agentic Development Security Platform unites security and development teams with actionable, code-to-runtime context to identify, prioritize, and fix the software risk that matters.

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

Cycode Workstation Protection intercepts package install commands on developer machines in real time, enforcing cooldown policies and blocking known malicious packages before installation. That closes a supply-chain control gap upstream of CI by applying policy at the workstation regardless of package manager version or local config.

Cycode added Workstation Protection to stop risky package installs directly on developer workstations before code enters the pipeline.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠ The release adds real-time install interception, cooldown policy enforcement, malicious package intelligence, and MDMMDMMobile device management, centrally configuring and managing enrolled devices. Administrators use it to apply settings, distribute applications and enforce device security policies. deployment without changing developer install workflows.

└──▷ WHAT SHIPPED · 3 FEATURES 3 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
Workstation package install enforcement new 75

Workstation Protection intercepts package install commands in real time on developer workstations to enforce cooldown, or release-age gating, policies and block known malicious packages before installation.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠ It checks every install against a continuously updated threat intelligence feed, including typosquattingtyposquattingUsing a name close to a trusted domain or software package to catch spelling mistakes. Attackers use it to lure people or installers to a malicious lookalike., slopsquattingslopsquattingPublishing malicious packages under names invented by AI coding tools. A developer who installs a hallucinated dependency may then retrieve the attacker's real package., and persistent compromised maintainer releases, and enforces policies at the device level regardless of package manager version or local config correctness.

— Explains behavior and scope, but no commands or UI path Introducing Cycode Workstation Protection: A…
thinner coverage below
02 02
MDM deployment for workstation protection new 47

Workstation Protection is deployed via Mobile Device Management, with no new console, no new infrastructure, and no change to developer install workflows.

— Names deployment route, but lacks setup steps Introducing Cycode Workstation Protection: A…
03 03
Workstation signals in Context Intelligence Graph improved 40

Cycode extended ADLC Security coverage upstream to the developer workstation by unifying workstation package enforcement with existing AI 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., prompt guardrails, pipeline integrity, 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. into the Context Intelligence GraphIntelligence GraphA graph database structure linking entities such as hosts, IPs, domains, and threat actors by relationship, letting cyber tools traverse connections to surface context that flat tables cannot express..

— Names connected areas, but not user-operable Introducing Cycode Workstation Protection: A…

Cycode Workstation Protection adds real-time package install enforcement with cooldown policies and malicious package threat intelligence.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠

  • Adds Workstation Protection, a new capability that intercepts package install commands in real time to enforce cooldown (release-age gating) policies and block known malicious packages before they install on developer workstations.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠
  • Deploys Workstation Protection via Mobile Device Management (MDMMDMMobile device management, centrally configuring and managing enrolled devices. Administrators use it to apply settings, distribute applications and enforce device security policies.) with no new console, no new infrastructure, and no change to developer install workflows.
  • Checks every package install against a continuously updated threat intelligence feed to catch malicious packages that have aged past cooldown windows, including typosquattingtyposquattingUsing a name close to a trusted domain or software package to catch spelling mistakes. Attackers use it to lure people or installers to a malicious lookalike., slopsquattingslopsquattingPublishing malicious packages under names invented by AI coding tools. A developer who installs a hallucinated dependency may then retrieve the attacker's real package., and persistent compromised maintainer releases.
  • Enforces cooldown policies at the device level, so gating applies regardless of which package manager version a developer or coding agent is running or whether a local config file was configured correctly.
  • Extends ADLC Security coverage upstream to the developer workstation, unifying workstation package enforcement with existing AI 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., prompt guardrails, pipeline integrity, 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. into the Context Intelligence GraphIntelligence GraphA graph database structure linking entities such as hosts, IPs, domains, and threat actors by relationship, letting cyber tools traverse connections to surface context that flat tables cannot express..
  • Adds Workstation Protection, a new capability that intercepts package install commands in real time to enforce cooldown (release-age gating) policies and block known malicious packages before they install on developer workstations.⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠⁠︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄⁠
  • Deploys Workstation Protection via Mobile Device Management (MDMMDMMobile device management, centrally configuring and managing enrolled devices. Administrators use it to apply settings, distribute applications and enforce device security policies.) with no new console, no new infrastructure, and no change to developer install workflows.
  • Checks every package install against a continuously updated threat intelligence feed to catch malicious packages that have aged past cooldown windows, including typosquattingtyposquattingUsing a name close to a trusted domain or software package to catch spelling mistakes. Attackers use it to lure people or installers to a malicious lookalike., slopsquattingslopsquattingPublishing malicious packages under names invented by AI coding tools. A developer who installs a hallucinated dependency may then retrieve the attacker's real package., and persistent compromised maintainer releases.
  • Enforces cooldown policies at the device level, so gating applies regardless of which package manager version a developer or coding agent is running or whether a local config file was configured correctly.

1 more in Everything

Good pick?
Good reason?

Where each pick stands · Sep 18–24

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

Each pick's release rhythm · Sep 18–24

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 18–24

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

  • Code Security 126usual 93.8
  • Identity & Access Management 104usual 74.2
  • SIEM 55usual 29.8
  • Data Security & DLP 40usual 11.5
  • Secure Access (VPN / SASE / ZTNA) 33usual 15.5
  • Secrets Detection & Management 32usual 13.2
  • Endpoint Security (EDR/XDR) 28usual 15.2
  • IDS / IPS 28usual 1.8
  • General Utilities & Libraries 23usual 15.8
  • Firewalls & Network Defense 22usual 9.5
  • Web Application Security 22usual 32.5
  • MCP Servers & Tooling 18usual 1

Who shipped the most · Sep 18–24

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

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