[Daily Report] Documentation Quality Report — April 19, 2026

[github-actions]

Daily status report on documentation quality for the gh-aw project, cross-referenced against open issues.

---

🆕 What Changed Since April 18

• workqueue-ops.md + batch-ops.md added (Apr 18, commit f7e3992): Both WorkQueueOps and BatchOps pattern pages are now live — resolves Add WorkQueueOps and BatchOps doc pages under design patterns #22001.
• pkg/constants spec drift fixed (issue Specification Audit — 2026-04-18 — 1 issue found: API drift in pkg/constants #27110, closed Apr 19): All 6 undocumented exported constants in pkg/constants/README.md patched by pelikhan.
• Copilot CLI 1.0.32 ([ca] CLI Version Updates: Claude Code 2.1.114, Copilot CLI 1.0.32 #27143, filed today): 11 new versions skipped; adds full COPILOT_PROVIDER_* BYOK env vars, --add-github-mcp-tool/toolset, hooks system, and 13 new model catalog entries — docs not yet updated.

---

🐛 Open Documentation Issues

1. #16498 — Provider-based authentication for Claude, Copilot, Codex (open since Feb 2026)

Labels: documentation, enhancement, security, cli

COPILOT_PROVIDER_* BYOK variables are fully implemented in CLI 1.0.32 (confirmed by #27143 today) but remain undocumented in user-facing docs. Zero occurrences of COPILOT_PROVIDER found anywhere in docs/.

Variables now available in CLI 1.0.32 that need documenting:

Variable	Purpose
COPILOT_PROVIDER_BASE_URL	API endpoint for custom model provider
COPILOT_PROVIDER_TYPE	Provider type: openai / azure / anthropic
COPILOT_PROVIDER_API_KEY	API key for custom provider
COPILOT_PROVIDER_BEARER_TOKEN	Bearer token for custom provider
COPILOT_PROVIDER_WIRE_API	API format: completions / responses
COPILOT_PROVIDER_MODEL_ID	Well-known model ID for agent config
COPILOT_PROVIDER_MAX_PROMPT_TOKENS	Max prompt tokens for BYOK model
COPILOT_PROVIDER_MAX_OUTPUT_TOKENS	Max output tokens for BYOK model

Relevant docs: reference/auth.mdx, reference/engines.md, reference/environment-variables.md
Status: ❌ Critical gap — BYOK is now fully shipped but completely undocumented. Open 2 months.

---

2. #20391 — Blog posts point to stale v0.45.5 install links (open since March 2026)

Labels: documentation, automation, workflows

115 occurrences of v0.45.5 found in the docs repo. Latest release is v0.68.3. Example stale link:

gh aw add-wizard https://github.com/github/gh-aw/blob/v0.45.5/.github/workflows/daily-doc-updater.md

Affects ~18 "Meet the Workflows" blog posts. No link-update automation exists.

Status: ❌ Ongoing — 115 stale links in blog posts; gap has been open 6 weeks.

---

3. #25383 — Add Playwright smoke test for editor playground in docs CI (open since April 8)

Labels: enhancement

The /editor/ playground broke silently once before due to CDN version drift. No Playwright test covers it; docs CI builds but never runs npm run test. Risk of silent regression remains.

Relevant docs: Testing infrastructure only — no doc content to fix.
Status: ❌ Open — no editor smoke test wired into docs CI.

---

✅ Issues Where Documentation Now Has the Answer

#22001 — Add WorkQueueOps and BatchOps doc pages (open since March 20)

Labels: documentation

Both pages were created on April 18 (commit f7e3992):

• /patterns/workqueue-ops/ — 4 queue strategies (issue checklist, sub-issues, cache-memory, discussions)
• /patterns/batch-ops/ — chunked processing, matrix fan-out, rate-limit-aware batching

Recommendation: This issue can be closed with a pointer to the new pattern pages.

---

🆕 New Issues With Documentation Implications

#27143 — Copilot CLI 1.0.32 (filed today)

Labels: automation, dependencies

Copilot CLI jumped 11 versions (1.0.21 → 1.0.32). Beyond the COPILOT_PROVIDER_* BYOK variables (tracked in #16498), the following new features also have no docs coverage:

• --add-github-mcp-tool / --add-github-mcp-toolset — granular GitHub MCP server tool control. Relevant to reference/tools.md and reference/gh-aw-as-mcp-server.md.
• hooks config key — inline hook definitions by event. New concept not mentioned anywhere.
• COPILOT_OFFLINE — offline mode env var. May be relevant to sandbox docs.
• 13 new model catalog entries (claude-opus-4.6/4.7, gpt-5.x series). reference/engines.md model list may be stale.

Status: 🆕 Documentation review needed for 1.0.32 feature set.

#27157 — Issue Monster: assign_to_agent field name bug (filed today)

Labels: automation, agentic-workflows

The issue-monster workflow emits issue-number (hyphen) but the schema requires issue_number (underscore), causing all agent assignments to fail silently. This is a workflow prompt bug, not a documentation bug — the spec at reference/safe-outputs-specification.md and safe_outputs_tools.json correctly use issue_number. However:

Documentation note: The assign_to_agent section in the spec (line 3061) lacks an explicit field schema table showing required fields. Adding issue_number / pull_number to a fields table would prevent this class of error.

Status: 🆕 Workflow fix in progress; minor spec improvement opportunity.

---

📊 Summary Table

Issue	Type	Docs Exist?	Action Needed
#16498 BYOK auth for Copilot	Missing content	❌ None	Add COPILOT_PROVIDER_* to reference/auth.mdx + engines.md
#20391 Stale v0.45.5 blog links	Stale links	❌ Stale (115 links)	Automate version link updates in blog
#22001 WorkQueueOps/BatchOps	✅ Added Apr 18	✅ patterns/workqueue-ops, patterns/batch-ops	Close issue
#25383 Editor playground CI	Testing gap	➖ N/A	Add Playwright test + wire docs CI
#27143 CLI 1.0.32 new features	New features	❌ Partial	Review hooks, MCP tool flags, offline mode
#27157 assign_to_agent field	Workflow bug	✅ Spec correct	Add field table to assign_to_agent spec entry

---

Report generated 2026-04-19 · workflow run

Note

🔒 Integrity filter blocked 17 items

The following items were blocked because they don't meet the GitHub integrity level.

• #20593 search_issues: has lower integrity than agent requires. The agent cannot read data with integrity below "approved".
• #25993 search_issues: has lower integrity than agent requires. The agent cannot read data with integrity below "approved".
• #26069 search_issues: has lower integrity than agent requires. The agent cannot read data with integrity below "approved".
• #18204 search_issues: has lower integrity than agent requires. The agent cannot read data with integrity below "approved".
• #19611 search_issues: has lower integrity than agent requires. The agent cannot read data with integrity below "approved".
• #19080 search_issues: has lower integrity than agent requires. The agent cannot read data with integrity below "approved".
• #21306 search_issues: has lower integrity than agent requires. The agent cannot read data with integrity below "approved".
• #26719 search_issues: has lower integrity than agent requires. The agent cannot read data with integrity below "approved".
• #17352 search_issues: has lower integrity than agent requires. The agent cannot read data with integrity below "approved".
• #22533 search_issues: has lower integrity than agent requires. The agent cannot read data with integrity below "approved".
• #17969 search_issues: has lower integrity than agent requires. The agent cannot read data with integrity below "approved".
• #21403 search_issues: has lower integrity than agent requires. The agent cannot read data with integrity below "approved".
• #24961 search_issues: has lower integrity than agent requires. The agent cannot read data with integrity below "approved".
• #18992 search_issues: has lower integrity than agent requires. The agent cannot read data with integrity below "approved".
• #21784 search_issues: has lower integrity than agent requires. The agent cannot read data with integrity below "approved".
• #21813 search_issues: has lower integrity than agent requires. The agent cannot read data with integrity below "approved".
• ... and 1 more item

To allow these resources, lower min-integrity in your GitHub frontmatter:

tools:
  github:
    min-integrity: approved  # merged | approved | unapproved | none

Generated by Dev · ● 1.9M · ◷

• expires on Apr 26, 2026, 9:07 AM UTC