Commit list:
test: add t.Parallel() to TestCompileCommandShortFlags ([test-parallel] Add t.Parallel() to TestCompileCommandShortFlags #49499 ) (680b762 )
fix: propagate model config into external threat detection path (fix: propagate model config into external threat detection path #49454 ) (ea271b2 )
Stabilize go-logger workflow toolchain resolution (Stabilize go-logger workflow toolchain resolution #49473 ) (2b2740f )
[evoskill] Refine developer skill with cross-trigger nullability guards for generated if: expressions ([evoskill] Refine developer skill with cross-trigger nullability guards for generated if: expressions #49472 ) (cb2a454 )
[code-simplifier] refactor: dedupe mention-alias normalization and JSONL iteration helpers in safe-output scripts ([code-simplifier] refactor: dedupe mention-alias normalization and JSONL iteration helpers in safe-output scripts #49462 ) (2248c79 )
Support explicit auto-merge strategies in safe-outputs.create-pull-request (Support explicit auto-merge strategies in safe-outputs.create-pull-request #49412 ) (25269b5 )
Fix Copilot BYOK target resolution for self-hosted provider URLs (Fix Copilot BYOK target resolution for self-hosted provider URLs #49464 ) (7ad1e24 )
Promote five custom analyzers to enforced linter gates (Promote five custom analyzers to enforced linter gates #49471 ) (6069546 )
Harden PR Description Updater against one-shot safe-output exhaustion (Harden PR Description Updater against one-shot safe-output exhaustion #49463 ) (d1adb68 )
fix: set GOTOOLCHAIN=auto in Avenger and hourly-ci-cleaner workflows (fix: set GOTOOLCHAIN=auto in Avenger and hourly-ci-cleaner workflows #49461 ) (f125eeb )
fix(daily-firewall-report): add missing create_discussion safe-output (fix(daily-firewall-report): add missing create_discussion safe-output #49460 ) (26f9ff2 )
dispatch_workflow: support per-call ref override for runtime branch targeting (dispatch_workflow: support per-call ref override for runtime branch targeting #49408 ) (a9137e1 )
[community] Update community contributions in README ([community] Update community contributions in README #49455 ) (55daba1 )
fix: extract sandbox.agent.memory from frontmatter so --memory-limit reaches AWF (fix: extract sandbox.agent.memory from frontmatter so --memory-limit reaches AWF #49448 ) (0c1d649 )
Extend stacked PR max-stack gating to pull_request_review (Extend stacked PR max-stack gating to pull_request_review #49453 ) (b5ab73e )
[model-inventory] Fix pricing for gpt-5.6-luna, gpt-5.6-terra, and grok-4.5 ([model-inventory] Fix pricing for gpt-5.6-luna, gpt-5.6-terra, and grok-4.5 #49449 ) (6dd674a )
Improve PR failure issue title specificity for engine HTTP 429 rate limits (Improve PR failure issue title specificity for engine HTTP 429 rate limits #49411 ) (d2d1e24)
Default stacked PR runs to top-of-stack with configurable on.pull_request.max-stack (Default stacked PR runs to top-of-stack with configurable on.pull_request.max-stack #49420 ) (184c9d1 )
Honor GH_AW_DETECTION_CONTINUE_ON_ERROR in detection setup to prevent false job failures (Honor GH_AW_DETECTION_CONTINUE_ON_ERROR in detection setup to prevent false job failures #49415 ) (ce344f3 )
docs: unbloat debug-ghe guide ([docs] docs: unbloat GHE data residency debugging guide #49442 ) (4a8907d )
chore: remove dead functions ([dead-code] chore: remove dead functions — 4 functions removed #49357 ) (465bf7d )
Handle persona-explorer invocation failures as first-class partial results (Handle persona-explorer invocation failures as first-class partial results #49413 ) (aceb490 )
Fix CLI consistency checker pre-agent help bootstrap to fail fast on empty datasets (Fix CLI consistency checker pre-agent help bootstrap to fail fast on empty datasets #49414 ) (2831507 )
[aw] Unblock Slide Deck Maintainer by enabling LFS checkout and shell command allowlist coverage ([aw] Unblock Slide Deck Maintainer by enabling LFS checkout and shell command allowlist coverage #49410 ) (113381f )
Prevent Super Linter false-negative failures by disabling broken summary formatting path (Prevent Super Linter false-negative failures by disabling broken summary formatting path #49409 ) (cf6e8d6 )
spdd 2026-07-31: cancelled fixture, safeguard TTL, copilot test refs, CTR audit (spdd 2026-07-31: cancelled fixture, safeguard TTL, copilot test refs, CTR audit #49379 ) (a4a8ff8 )
Fix AI credits throughput rate-limit errors being silently swallowed (Fix AI credits throughput rate-limit errors being silently swallowed #49360 ) (498f17f )
Fix docs build to prerender llms.txt (Docs: prerender llms.txt so GitHub Pages publishes /llms.txt #49353 ) (61a96fb )
Fix gh aw logs discarding all processed runs due to variable shadowing (Fix gh aw logs discarding all processed runs due to variable shadowing #49332 ) (ccf88b8 )
Fix evals judge parser dropping YES/NO answers as UNKNOWN for Claude engine output (Fix evals judge parser dropping YES/NO answers as UNKNOWN for Claude engine output #49323 ) (b0541d0 )
Summary:
Recent commits focus on workflow and safe-output reliability, with several fixes around PR handling, workflow dispatching, and detection/CI behavior. There are also model/configuration updates, including BYOK target resolution and pricing/inventory changes, along with documentation and community updates. The overall area is platform robustness and developer experience improvements.
Bash tool test: ✅
Overall status: PASS
Run URL: https://github.com/github/gh-aw/actions/runs/30689986700
Timestamp: 2026-08-01T07:36:19Z
Generated by ⚡ Smoke Copilot MAI · mai10 · 3.1 AIC · ⌖ 1.51 AIC · ⊞ 19.7K · ◷
Comment /smoke-copilot-mai to run again
Add label smoke to run again
Commit list:
if:expressions ([evoskill] Refine developer skill with cross-trigger nullability guards for generatedif:expressions #49472) (cb2a454)safe-outputs.create-pull-request(Support explicit auto-merge strategies insafe-outputs.create-pull-request#49412) (25269b5)refoverride for runtime branch targeting (dispatch_workflow: support per-callrefoverride for runtime branch targeting #49408) (a9137e1)max-stackgating topull_request_review(Extend stacked PRmax-stackgating topull_request_review#49453) (b5ab73e)on.pull_request.max-stack(Default stacked PR runs to top-of-stack with configurableon.pull_request.max-stack#49420) (184c9d1)llms.txtso GitHub Pages publishes/llms.txt#49353) (61a96fb)gh aw logsdiscarding all processed runs due to variable shadowing (Fixgh aw logsdiscarding all processed runs due to variable shadowing #49332) (ccf88b8)Summary:
Recent commits focus on workflow and safe-output reliability, with several fixes around PR handling, workflow dispatching, and detection/CI behavior. There are also model/configuration updates, including BYOK target resolution and pricing/inventory changes, along with documentation and community updates. The overall area is platform robustness and developer experience improvements.
Bash tool test: ✅
Overall status: PASS
Run URL: https://github.com/github/gh-aw/actions/runs/30689986700
Timestamp: 2026-08-01T07:36:19Z