[instructions] Sync instruction files with release v0.88.7 (assign-to-agent reasoning-effort) - #60800
Conversation
Add the reasoning-effort field (added in PR #60346) to the assign-to-agent example in safe-outputs-automation.md; it was documented in docs/ but missing from the .github/aw instructions.
|
✅ Looks great! This PR adds the The audit in the PR body is thorough: you've verified all files remain under target line limits, confirmed field names match the implementation, and identified the documentation gap from #60346. This is exactly the kind of small, focused, well-justified change that keeps the instructions aligned with the codebase. Ready to merge! 🎯 Warning Firewall blocked 1 domainThe following domain was blocked by the firewall during workflow execution:
To allow these domains, add them to the network:
allowed:
- defaults
- "github.com"See Network Configuration for more information.
|
Summary
Ran the instructions-janitor audit against the latest release (v0.88.7, 2026-09-08) and the two prior instruction syncs (#60231, #60603 — most recent at 2026-09-13T14:36:23Z).
Release and change audit
Reviewed all commits touching
docs/and.github/aw/since the release, focusing on what postdates the most recent sync (#60603):Require update-release for agent release notes(Require update-release for agent-generated release notes #60701) — already updated.github/aw/*directly in the same commit; no action needed.Bump gh-aw-firewall/gh-aw-mcpgversion bump (Bump gh-aw-firewall to v0.28.16 and gh-aw-mcpg to v0.4.21 #60568) — only touched.github/aw/actions-lock.jsonand workflow lock files; not instruction content.docs: unbloat frontmatter reference([docs] docs: unbloat frontmatter reference #60713),[docs] Correct artifact names in artifacts reference([docs] Self-healing documentation fixes from issue analysis - 2026-09-13 #60706),Support wildcard cached logs files(Support wildcard cached logs files #60702) — all scoped todocs/(website reference pages);.github/aw/cli-commands.md'sgh aw logssection is intentionally a compact, non-exhaustive flag list and doesn't need the new--cached-logsalias.assign-to-agentreasoning effort,add-labelstarget restrictions, OTLP secret-backed endpoints, safe-output footers) to confirm they were captured by the prior sync. One gap surfaced:reasoning-effortonassign-to-agent(Add reasoning effort support to assign-to-agent #60346) was documented indocs/but missing from.github/aw/safe-outputs-automation.md's example block — fixed here.Size audit
wc -l .github/aw/*.md— all 68 files remain under their 400-line targets (largest:syntax-agentic.mdat 391,token-optimization.mdat 384,safe-outputs-content.mdat 379,create-agentic-workflow.mdat 376). No split needed this run.Duplication / accuracy audit
No new duplication introduced by recent changes;
add-labelstarget restrictions (#60362) and OTLP secret-backed defaults (#60348) were internal/enforcement changes with no new user-facing frontmatter fields, so no doc changes were needed there.Files changed
.github/aw/safe-outputs-automation.mdTest plan
wc -l .github/aw/*.md— confirmed all files under target limitsreasoning-effortmatches the field name/behavior inpkg/workflow/assign_to_agent.goanddocs/src/content/docs/reference/safe-outputs.mdWarning
Firewall blocked 1 domain
The following domain was blocked by the firewall during workflow execution:
api.anthropic.comTo allow these domains, add them to the
network.allowedlist in your workflow frontmatter:See Network Configuration for more information.