[docs] docs: unbloat maintaining-repos guide

[github-actions]

Reduced docs/src/content/docs/guides/maintaining-repos.md from 512 to 399 lines — a 22% reduction — while preserving all technical information.

Changes Made

1. Common Failure Patterns → table (58 lines → 9 lines)

The six failure patterns each had the same prose structure: bolded heading, description sentence, then a "Fixes:" bullet list. Converted to a single 3-column table (Failure / Symptom–Cause / Fixes), keeping every fix and link intact.

2. Fine-Grained Trust Controls → consolidated code block (30 lines → 18 lines)

trusted-users, approval-labels, and blocked-users each had their own identical tools.github wrapper code block. Combined into one block showing all three fields together, and converted the prose introductions to a compact bullet list.

3. Public OSS Worked Example → summary paragraph (54 lines → 4 lines)

The "Public Open-Source Repository" worked example was near-identical to the two-agent pattern shown in the document's introduction (same workflow structure, same fields, only the label name differed). Replaced both full workflow code blocks with a one-paragraph description referencing the intro pattern and noting the agent-ready label difference.

Stats

Metric	Before	After
Lines	512	399
Reduction	—	22%
Bullet points in failure section	18	0 (→ table)
Duplicate code blocks	3	1

Screenshots

Visual screenshots were unavailable — the Playwright browser container could not reach the agent container's bridge IP (net::ERR_CONNECTION_TIMED_OUT). The page was verified via curl to return HTTP 200 with the correct title ("Maintaining Repos with Agentic Workflows").

References: §24596068735

🗜️ Compressed by Documentation Unbloat · ● 243.5K · ◷

• expires on Apr 20, 2026, 3:47 AM UTC

[Copilot]

Pull request overview

This PR reduces verbosity in the “Maintaining Repos with Agentic Workflows” guide while keeping the guidance focused and easier to scan.

Changes:

• Consolidates “Fine-Grained Trust Controls” into a single tools.github example and a compact bullet list.
• Reformats “Common Failure Patterns” from repeated prose blocks into a single table.
• Replaces the “Public Open-Source Repository” worked example’s duplicated workflow blocks with a summary paragraph referencing earlier patterns.

Show a summary per file

File	Description
docs/src/content/docs/guides/maintaining-repos.md	Compresses repeated docs content via tables/merged examples and removes duplicated workflow blocks.

Copilot's findings

Tip

Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

• Files reviewed: 1/1 changed files
• Comments generated: 2

[Copilot]

Link text "Auth Reference" is inconsistent with the rest of the docs (this guide previously used "Authentication Reference", and many other pages use "Authentication" / "Authentication Reference" for the same URL). Consider renaming the link text back to "Authentication Reference" for consistency and easier scanning.

Suggested change

	| **Authentication failures** | Token permissions too narrow or API key missing. | Review `permissions:` block; ensure secrets are set; see [Auth Reference](/gh-aw/reference/auth/). |
	| **Authentication failures** | Token permissions too narrow or API key missing. | Review `permissions:` block; ensure secrets are set; see [Authentication Reference](/gh-aw/reference/auth/). |

[Copilot]

This paragraph references the "two-agent pattern shown above", but the earlier worked example uses the label needs-investigation while this section uses agent-ready. To avoid confusion for readers copying the earlier config, consider explicitly stating that this is the same pattern as the earlier "Routing to Downstream Agents" example but with the approval label renamed to agent-ready (and/or link to that section).

Suggested change

A public repository uses the two-agent pattern shown above: repo-assist with `min-integrity: unapproved` classifies incoming issues and applies an `agent-ready` label to quality bug reports, and a downstream code fix agent with `min-integrity: approved` and `approval-labels: ["agent-ready"]` opens a pull request only after repo-assist has promoted the issue. Issues from untrusted users can still trigger the pipeline through label promotion, while the code fix agent never sees unapproved content directly.

A public repository uses the same two-agent routing pattern as the earlier "Routing to Downstream Agents" example, but renames the approval label from `needs-investigation` to `agent-ready`: repo-assist with `min-integrity: unapproved` classifies incoming issues and applies an `agent-ready` label to quality bug reports, and a downstream code fix agent with `min-integrity: approved` and `approval-labels: ["agent-ready"]` opens a pull request only after repo-assist has promoted the issue. Issues from untrusted users can still trigger the pipeline through label promotion, while the code fix agent never sees unapproved content directly.