Skip to content

chore(deps): update dependency github/gh-aw to v0.89.21 - #401

Open
bootc-bot[bot] wants to merge 1 commit into
mainfrom
bootc-renovate/github-actions
Open

chore(deps): update dependency github/gh-aw to v0.89.21#401
bootc-bot[bot] wants to merge 1 commit into
mainfrom
bootc-renovate/github-actions

Conversation

@bootc-bot

@bootc-bot bootc-bot Bot commented Oct 4, 2026 •
edited
Loading

Copy link
Copy Markdown
Contributor

ℹ️ Note

This PR body was truncated due to platform limits.

This PR contains the following updates:

Package Type Update Change
github/gh-aw uses-with minor v0.88.7 → v0.89.21

Release Notes

github/gh-aw

v0.89.21

Compare Source

🌟 Release Highlights

This release brings native web-search support for the Copilot engine, more flexible reusable-workflow failure reporting, and hardened safe-outputs checkout detection.

✨ What's New
  • Native web-search on the Copilot engine — tools: web-search: now compiles to Copilot's built-in web_search tool (--allow-tool web_search) instead of producing a compile warning, making web search usable even in repos without GitHub tooling (e.g. Azure DevOps-hosted). See Web Search reference. (#​62957)
  • Dynamic failure-issue-repo for reusable workflows — safe-outputs.failure-issue-repo now accepts ${{ inputs.* }} expressions, matching existing support for report-failure-as-issue and report-failed-jobs, so reusable (workflow_call) workflows can route failure issues per caller without patching the compiled lock file. (#​62945)
  • Gateway steering events in audit output — gh aw audit now surfaces token_steering and timeout_steering events (type, message, timestamp) as gateway_steering_events in both JSON and console output, making it easier to see when runs are approaching AI Credit or time limits. (#​62943)
🐛 Bug Fixes & Improvements
  • Fixed cancelled AIC component accounting — cancelled compiler-owned component jobs are now only counted as zero AI Credits when GitHub job metadata proves execution never started; jobs with assigned runners, steps, or incomplete metadata are still accounted for correctly. (#​62984)
  • Safe-outputs checkout detection fixed for nested repos — the find_repo_checkout git-scan fallback now trusts scanned repositories cloned via a steps: entry or manual actions/checkout, fixing a regression where the containerized safe-outputs MCP server (different UID) couldn't read nested checkouts due to git's safe.directory trust not propagating beyond GITHUB_WORKSPACE. (#​62944)
📚 Documentation
  • Stale lock-file detection guidance — new docs explain how to detect stale/missing compiled .lock.yml files via gh aw list --json for CI enforcement, plus remediation steps for contributors. (#​62947)
  • Portable OTLP helper paths — OpenTelemetry/qmd examples now resolve the helper via RUNNER_TEMP instead of a hardcoded /tmp, for portability across runner environments. (#​62940)
🔧 Internal
  • Updated GitHub Actions versions and refreshed parser/repoutil/semverutil/sliceutil spec extractions.

[!WARNING]

Firewall blocked 1 domain

The following domain was blocked by the firewall during workflow execution:

  • o205451.ingest.us.sentry.io

To allow these domains, add them to the network.allowed list in your workflow frontmatter:

network:
  allowed:
    - defaults
    - "o205451.ingest.us.sentry.io"

See Network Configuration for more information.

Generated by 🚀 Release · copilot · auto · 21.7 AIC · ⊞ 11.6K

What's Changed

Full Changelog: github/gh-aw@v0.89.20...v0.89.21

v0.89.20

Compare Source

🌟 Release Highlights

This release focuses on reliability fixes for cross-repo pull request handling, threat-detection infrastructure, and Linear/Jira safe-output integrations, alongside continued documentation and automation upkeep.

🐛 Bug Fixes & Improvements
  • Fixed cross-org fork PR creation — create_pull_request now passes the validated head repository, so PRs from forks that share an organization with the base repo no longer fail (#​62720).
  • Fixed external threat detection on ARC/DinD runners — the detector binary and inputs are now staged under the shared RUNNER_TEMP directory and installed rootlessly, resolving failures where the runner and Docker daemon don't share a filesystem (#​62755).
  • Hardened Linear integration — create_issue now accepts a team UUID, key, or friendly name instead of requiring a UUID (#​62754), and workflow activation now validates LINEAR_TEAM_ID up front to prevent silent handler load failures (#​62743).
  • Improved threat-detection auditing — gh aw audit now surfaces failed detection jobs and reports detected threats with stable finding codes for downstream automation (#​62753).
  • Fixed installer robustness — gh-aw binary downloads are now staged before replacing existing binaries, preventing partial upgrades (#​62761).
  • Retry logic added for transient workflows-scope timeouts when pushing branches for create_pull_request and push_to_pull_request_branch (#​62486, #​62668).
  • Allowed line-oriented inspection tools (nl, sed) in the PR Code Quality Reviewer to prevent premature harness stops (#​62770).
✨ What's New
  • Jira safe outputs promoted to stable — jira-create-issue, jira-update-issue, jira-add-comment, and jira-add-label are now stable, production-ready safe outputs (#​62749).
📚 Documentation
  • Refreshed the repo assist example and Playwright reference for clarity (#​62822, #​62277).
  • Routine Astro dependency and glossary updates keep the docs site current (#​62853, #​62610).

[!WARNING]

Firewall blocked 1 domain

The following domain was blocked by the firewall during workflow execution:

  • o205451.ingest.us.sentry.io

To allow these domains, add them to the network.allowed list in your workflow frontmatter:

network:
  allowed:
    - defaults
    - "o205451.ingest.us.sentry.io"

See Network Configuration for more information.

Generated by 🚀 Release · copilot · auto · 20 AIC · ⊞ 11.6K

What's Changed

Full Changelog: github/gh-aw@v0.89.19...v0.89.20

v0.89.17

Compare Source

🌟 Release Highlights

This release focuses on hardening reliability across the AIC accounting pipeline, AWF/firewall integration, and safe-outputs handling, alongside a refreshed model catalog and several documentation clean-ups.

✨ What's New
  • Faster, smarter logs auditing — cached workflow runs are no longer redownloaded during logs audits (#​61871), and multi-target logs queries are now distributed fairly across targets (#​61027), with per-run download duration/size tracked in an end-of-run stats summary (#​60951).
  • Updated model catalog — added gemini-3.8-flash and claude-fable-5.1 aliases and corrected pricing for gpt-6-astra/gpt-5.6-sol (#​61234).
  • MCP Gateway and firewall bumped — MCP Gateway updated to v0.4.25 (#​61661) and gh-aw-firewall (AWF) updated to v0.28.20 (#​61527) and v0.28.17 (#​60945), improving compatibility and stability.
  • Better automatic grading — native Copilot tool calls are now included in the automatic grader trace payload for more accurate evaluation (#​61426).
  • Refreshed CLI defaults and scanner image pins for more predictable compiled workflows (#​61432).
🐛 Bug Fixes & Improvements
  • Fixed Code Scanning Fixer timeouts and tool denials (#​61605).
  • Fixed slash command activation failing on CRLF line endings (#​61602).
  • Added support for older curl versions in the AWF installer (#​61600).
  • Rewrote experiments.<name> references in engine.model into valid job-scoped expressions, preventing invalid compiled workflows (#​61599).
  • Hardened withRetry against transient fetch failures (#​61439).
  • Fixed upload_artifact silently succeeding when relative paths were never staged (#​61431).
  • Fixed Copilot SDK multiword shell-prefix matching and a denial-guard hang (#​61430).
  • Empty add_labels lists are now treated as a no-op instead of failing the job (#​61429).
  • Added ability to opt out of the "[aw] Detection Runs" tracking issue independently of threat detection (#​61428).
  • The safeoutputs CLI transport now fails loudly instead of silently failing open, surfacing real errors sooner (#​61427).
  • Fixed false-positive AI credits rate-limit detection caused by MCP echoes (#​61425).
  • Imported engine config (including auth) is now preserved when a workflow sets a top-level model (#​61424).
  • PLAYWRIGHT_BROWSERS_PATH now uses ${{ runner.temp }} so install and launch agree on the browser path (#​61423).
  • Fixed several gaps in daily AIC (AI Credits) accounting: legacy runs (#​61313), pre-harness failures (#​61232), unassigned jobs (#​61222), and missing evals now counted as zero instead of skipped (#​60892).
  • Stabilized the Daily Documentation Healer's runtime (#​61235).
  • AWF fatal startup errors are now surfaced in agent failure reports (#​61199), and not_started execution evidence is recorded when AWF fails before the engine harness starts (#​61202).
  • Added the Go ecosystem to the network allowlist for ci-coach (#​61201).
  • Validated cached run uniqueness across JSONL shards to prevent duplicate accounting (#​61220).
  • Preserved agent accounting in fallback artifacts (#​61053).
  • Fixed a flaky pkg/cli test-unit crash from concurrent Cobra completion generation (#​61146).
📚 Documentation
  • Several self-healing documentation passes fixed inaccuracies found via issue analysis (#​61683, #​61443).
  • Documented the threat-detection report-as-issue field (#​61563).
  • Trimmed and clarified the workflow structure, tools, and IssueOps reference docs (#​61485, #​61238, #​60998).
  • Updated the glossary from the daily scan (#​61104).

Generated by 🚀 Release · copilot · auto · 18.5 AIC · ⊞ 11.6K

What's Changed

Full Changelog: github/gh-aw@v0.89.16...v0.89.17

v0.89.15

Compare Source

What's Changed

Full Changelog: github/gh-aw@v0.89.14...v0.89.15

v0.89.13

Compare Source

🌟 Release Highlights

A small maintenance release focused on logs tooling, release-notes governance, and a documentation fix.

✨ What's New
  • Wildcard support for cached logs — gh aw logs --cached-logs now accepts trailing-wildcard prefixes (e.g. logs-*), merging all matching *.jsonl shards, pruning out-of-range files, and writing fresh results to a new collision-resistant shard. Handy for teams sharing rolling log caches. (#​60702)
  • Safer release-notes automation — agentic workflows that touch release notes must now go through the update-release safe output instead of direct GitHub mutations, keeping agent jobs read-only and closing off a class of unintended writes. (#​60701)
📚 Documentation
  • Fixed the Artifacts reference table, which listed aw-info and prompt as current artifact names even though no compiled workflow produces them; the correct info artifact is now documented, and the legacy names are marked back-compat. (#​60706)
🔧 Internal
  • Weekly blog post summarizing recent activity. (#​60729)

No community-labeled issues were closed in this release window.

Generated by 🚀 Release · copilot · auto · 23.1 AIC · ⊞ 11.9K

What's Changed

Full Changelog: github/gh-aw@v0.89.12...v0.89.13

v0.89.12

Compare Source

🌟 Release Highlights

A small security-hardening and CI reliability release.

🐛 Bug Fixes & Improvements
  • Reduced credential blast radius in slash-command router — The generated central slash-command router workflow (.github/workflows/agentic_commands.yml) now checks out the repository with persist-credentials: false, so GITHUB_TOKEN is no longer persisted in the local git config for the lifetime of the routing job (#​60685).
  • Fixed failing CI job — Resolved a failure in the "Integration: CMD Tests" GitHub Actions job to restore green CI signal (#​60683).

Generated by 🚀 Release · copilot · auto · 19.3 AIC · ⊞ 11.9K

What's Changed

Full Changelog: github/gh-aw@v0.89.11...v0.89.12

v0.89.11

Compare Source

🌟 Release Highlights

This release focuses on reliability hardening for agentic workflow execution — fixing model-compatibility bugs, improving error detection, and closing safety gaps in the linter autofix pipeline.

🐛 Bug Fixes & Improvements
  • Issue Monster model fixes (#​60561, #​60661): Resolved Pi engine model-routing failures that prevented Issue Monster workflows from starting, pinning it to a compatible model configuration.
  • Guard linter autofixes from comment-overlap deletions (#​60559): Closed a gap where 13 autofix builders could silently drop inline comments during -fix runs; added regression coverage to keep diagnostics safe.
  • Detect standalone Copilot CLI HTTP 400 failures (#​60647): The harness now correctly classifies standalone 400 Bad Request responses instead of misreporting them as partial execution, avoiding futile retries.
  • Handle negative API rate limits across log targets (#​60636): Multi-target log collection now correctly treats a negative GitHub API limit as a shared reserve, stopping all targets once the ceiling is reached.
  • Preserve JSONL rows during repo-memory merge conflicts (#​60663): Concurrent repo-memory pushes now use a union merge policy for .jsonl files so rows added by parallel workflow runs are no longer discarded.
  • Fix known broken evals accounting in daily AIC guardrail (#​60648): The guardrail now recognizes a verified zero-usage run case while preserving fail-closed behavior for unrecognized failures.
  • Disable persisted credentials in auto-upgrade checkout (#​60650): The generated auto-upgrade workflow no longer retains repository credentials after checkout.
🔧 Dependencies
  • Bumped gh-aw-firewall to v0.28.16 and gh-aw-mcpg to v0.4.21 (#​60568), refreshing container digest pins across all workflow lock files.
🤖 New Linter
  • Added a new Go analyzer, bufio-scanner-err-unchecked (#​60660), which flags bufio.Scanner loops that don't check Err() after completion — catching silently dropped I/O errors. Found and helped fix 9 real instances in this codebase.

No community-labeled issues were resolved in this release window.

Generated by 🚀 Release · copilot · auto · 23.6 AIC · ⊞ 11.9K

What's Changed

Full Changelog: github/gh-aw@v0.89.10...v0.89.11

v0.89.10

Compare Source

🌟 Release Highlights

This release focuses on reliability improvements to the AI Credits (AIC) accounting guardrail, plus a package-manifest enhancement for shared JavaScript workflow modules.

✨ What's New
  • Shared JavaScript modules in aw.yml packages — Package manifests can now ship .mjs and .cjs helper files alongside shared workflows under .github/workflows/shared/, with path traversal and extension restrictions preserved. See the imports reference.
  • Coordinated rate-limit handling for multi-target gh aw logs — Rate-limit state is now shared across concurrent log-download targets, so queued work is cancelled once the shared ceiling is reached instead of blocking until reset. Single-target downloads retain the existing wait-for-reset behavior.
🐛 Bug Fixes & Improvements
  • More accurate AI Credits accounting for failed runs — The daily AIC guardrail now correctly treats failed agent runs and evals with genuinely zero token usage as zero cost, instead of failing closed on unknown usage. This applies to:
    • Failed agent runs with an empty (but successfully collected) usage accounting file.
    • Evals jobs that failed before model execution ever started, using new execution-state evidence (not_started → started) to distinguish "never ran" from "lost accounting data."
    • Evals jobs whose token collector completed successfully but reported zero usage.
    • In all cases, the guardrail continues to fail closed whenever token collection itself fails or doesn't run, preserving safety.
📚 Documentation
  • Condensed and clarified the imports reference documentation, tightening explanations of shared components, path resolution, and merge strategies without changing behavior.

Generated by 🚀 Release · copilot · auto · 20.7 AIC · ⊞ 11.9K

What's Changed

Full Changelog: github/gh-aw@v0.89.9...v0.89.10

v0.89.9

Compare Source

🌟 Release Highlights

This release focuses on reliability improvements for AI credit accounting, logs tooling, and workflow updates, plus a couple of targeted bug fixes.

✨ What's New
  • Package-aware gh aw update — the update command now targets packages individually, making it easier to keep multi-package workflow setups current (#​60452).
  • Node network preset for Impeccable Skills Reviewer — adds a preconfigured network preset so the reviewer workflow can access Node-related resources out of the box (#​60425, closes #​60411). See the network configuration guide.
🐛 Bug Fixes & Improvements
  • Fixed Codex model identifier handling — Codex engine now correctly strips the openai/ provider prefix from model identifiers, preventing misconfigured model lookups (#​60423, closes #​60416).
  • Fixed zero token usage in logs MCP tool — corrected the run schema so token usage is now accurately reported (#​60424, closes #​60412).
  • Fixed Issue Monster's unsupported Pi model — resolves a compatibility issue that could break the Issue Monster workflow (#​60461).
  • Preserved zero-credit threat detection accounting — ensures threat detection runs with zero credit usage are still accurately accounted for (#​60380).
  • Improved daily AI Credits accounting diagnostics — clearer failure messages make it easier to troubleshoot accounting issues (#​60433).
  • Improved cached logs reporting — better visibility into cached vs. fresh log data (#​60430).
  • Added live integration test coverage for logs JSONL caching, improving confidence in future changes (#​60447).

Generated by 🚀 Release · copilot · auto · 20.4 AIC · ⊞ 11.9K

What's Changed

Full Changelog: github/gh-aw@v0.89.8...v0.89.9

v0.89.8

Compare Source

🌟 Release Highlights

This release strengthens billing and guardrail accuracy for AI Credits, hardens safe-output permission enforcement, and adds new configuration flexibility for OTLP telemetry and agent assignment.

✨ What's New
  • Secret-backed OTLP endpoints — Organization-wide OpenTelemetry endpoints can now be configured via secrets.GH_AW_DEFAULT_OTLP_ENDPOINT in addition to visible variables, with automatic fallback and existing precedence preserved.
  • Reasoning effort for assign-to-agent — Forward model-specific reasoning_effort (e.g. high) through agent assignment, with automatic validation and graceful fallback for unsupported agents/models.
  • continue-on-error on generated agent jobs — The compiler now correctly preserves an explicit continue-on-error: true/false on the generated agent job instead of silently dropping it.
  • Date-bounded cached JSONL logs — gh aw logs --cached-jsonl now prunes cached run records outside the requested --start-date/--end-date range instead of only appending, keeping cached history accurate.
  • Centralized slash commands get PR read access — pre_activation now automatically receives pull-requests: read when a workflow defines a centralized slash command, fixing silent execution skips on pull requests.
🐛 Bug Fixes & Improvements
  • Fixed the daily AI Credits gate incorrectly rejecting failed agent runs that had authoritative zero-credit usage artifacts.
  • Fixed the daily AIC guardrail treating pre-agent setup failures (with no model usage) as potentially billable, by tracking explicit execution-started/not-started evidence.
  • Hardened add_labels to enforce configured targets, preventing labels from being applied outside the triggering issue or PR unless explicitly allowed via target: "*".
  • Prevented benign tool-result text (e.g. "not logged in", EACCES) in Claude stream-JSON output from incorrectly triggering failure classifiers and suppressing retries.
📚 Documentation
  • Tightened the trace graders documentation for clarity and conciseness.

Generated by 🚀 Release · copilot · auto · 19.6 AIC · ⊞ 11.9K

What's Changed

Full Changelog: github/gh-aw@v0.89.7...v0.89.8

v0.89.7

Compare Source

🌟 Release Highlights

A focused maintenance release improving the reliability of gh aw logs multi-target downloads and tightening the safe-outputs compiler for enclave configurations.

✨ What's New
  • Faster, more predictable gh aw logs --count downloads — When multiple workflow targets share a --count budget, sibling targets are now cancelled promptly as soon as the shared limit is reached, instead of finishing an in-flight over-fetched chunk first (#​60323, #​60343). This reduces wasted API calls and speeds up multi-target log retrieval.
🐛 Bug Fixes & Improvements
  • Fixed a compiler bug where GH_AW_SINK_VISIBILITY referenced a determine-automatic-lockdown step that was never generated when tools.github: false was combined with a static GitHub enclave and safe-outputs, which could break workflow compilation for that configuration (#​60338).
  • Resolved a failing lint-go-golangci CI job (#​60316).
🔧 Internal
  • Added a new custom Go analysis linter, slice-make-zero-length, to catch make([]T, 0) calls without a capacity argument (#​60310).

No breaking changes in this release.

Generated by 🚀 Release · copilot · auto · 21.6 AIC · ⊞ 11.9K

What's Changed

Full Changelog: github/gh-aw@v0.89.6...v0.89.7

v0.89.6

Compare Source

🌟 Release Highlights

A small, focused patch release: a gh aw logs reliability fix for multi-target downloads and a compiler tweak that relaxes an experimental-runtime warning.

🐛 Bug Fixes & Improvements
  • gh aw logs multi-target downloads now respect --count/--timeout — Previously, running gh aw logs against multiple targets could over-fetch (~3× the requested count per target) and silently truncate results when a shared timeout expired mid-run, leaving continuations with timeout: 0. Each target now caps its work against the remaining shared budget, and per-target timeouts are correctly inherited so continuations replay with the original timeout. (#​60308)
  • Removed the mandatory security review warning for cloud-hypervisor sandbox runtime — The compiler no longer emits this warning when sandbox.agent.runtime: cloud-hypervisor is set; the runtime remains in experimental preview and still requires --cloud-hypervisor-preview. (#​60304)

No community-labeled issues were closed in this release window.

Generated by 🚀 Release · copilot · auto · 19.8 AIC · ⊞ 11.9K

What's Changed

Full Changelog: github/gh-aw@v0.89.5...v0.89.6

v0.89.5

Compare Source

🌟 Release Highlights

This release focuses on hardening safe-output reliability, tightening security around confused-deputy protections, and improving multi-target log collection under load.

✨ What's New
  • Configurable safe-output footers — safe-outputs.body-footer lets you append shared, additive content to every safe output that produces a body, composing cleanly across imported workflows and handler-specific config. Learn more
  • Early validation for external safe-output secrets — Jira, Linear, and Azure DevOps safe outputs now fail fast at activation time with actionable errors when required credentials are missing, instead of failing mid-run after agent execution.
  • GHE-compatible activation artifacts — Activation jobs now upload aw_info.json as an archived artifact, fixing compatibility with GitHub Enterprise where unarchived artifacts aren't supported.
🐛 Bug Fixes & Improvements
  • Fixed daily AIC snapshot accounting — Resolved reuse of incomplete observations, repeated re-fetching of historical artifacts, and inaccurate "under budget" reporting after quota failures.
  • Honored bot allowlists before confused-deputy checks — Allowlisted bots pushing to PRs opened by another actor were previously rejected on synchronize events; the bot allowlist is now checked first.
  • More reliable multi-target log collection — Queued log targets no longer start after the shared count limit or timeout has already been reached, with clearer diagnostic logging for cache reuse and active limits.
🔧 Internal
  • Updated GitHub Actions versions, dependency bumps (@primer/octicons, @types/node, @astrojs/sitemap, starlight-links-validator, @astrojs/starlight), package specification refreshes, and dead-code removal.

Generated by 🚀 Release · copilot · auto · 17.9 AIC · ⊞ 11.9K

What's Changed

❗ Important

✂ PR body was truncated to here.

Configuration

📅 Schedule: (in timezone UTC)

  • Branch creation
    • "on sunday"
  • Automerge
    • At any time (no schedule defined)

🚦 Automerge: Disabled by config. Please merge this manually once you are satisfied.

♻ Rebasing: Whenever PR becomes conflicted, or you tick the rebase/retry checkbox.

🔕 Ignore: Close this PR and you won't be reminded about this update again.

  • If you want to rebase/retry this PR, check this box

This PR has been generated by Mend Renovate CLI.

@bootc-bot
chore(deps): update dependency github/gh-aw to v0.89.21
03e2a43 
Signed-off-by: bootc-bot[bot] <225049296+bootc-bot[bot]@users.noreply.github.com>
Signed-off-by: bootc-bot[bot] <225049296+bootc-bot[bot]@users.noreply.github.com>
@bootc-bot
bootc-bot Bot force-pushed the bootc-renovate/github-actions branch from 8a1f1d7 to 03e2a43 Compare October 11, 2026 01:07
@bootc-bot bootc-bot Bot changed the title chore(deps): update github/gh-aw-actions action to v0.90.3 chore(deps): update dependency github/gh-aw to v0.89.21 Oct 11, 2026

This branch has not been deployed

No deployments
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

No reviews

Assignees

No one assigned

Labels

None yet

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

0 participants