Skip to content

blog: introduce Agent Shin triage bot for external PRs and issues#240

Open
mateo-berri wants to merge 1 commit into
mainfrom
mateo/agent-shin-triage-blog
Open

blog: introduce Agent Shin triage bot for external PRs and issues#240
mateo-berri wants to merge 1 commit into
mainfrom
mateo/agent-shin-triage-blog

Conversation

@mateo-berri
Copy link
Copy Markdown
Collaborator

@mateo-berri mateo-berri commented May 28, 2026

Summary

  • Adds blog/agent_shin_triage/index.md — the public-facing explainer linked from every Agent Shin comment on external PRs and issues.
  • Documents both rubrics: PR (context + non-mocked QA proof) and issue (bug-report vs. feature-request requirements).
  • Explains the 24h grace window, the @agent-shin reconsider / @greptileai recovery paths, and how a closed PR/issue can be reopened.

Renders at `docs.litellm.ai/blog/agent-shin-triage` (slug).

Known follow-ups (not in this PR)

  • The grace-warning + close comments in `.github/scripts/triage_with_llm.py` (over on `BerriAI/litellm`) still say "before/after screenshots, a short screen recording, or terminal/log output" without the no-mocking clarification. If we want full consistency, those two strings should be updated to match the post.
  • The "If you disagree" section links to the generic issues list — if we want a dedicated rubric-feedback thread, file one and I'll swap the URL.

Test plan

  • Local preview renders the post with correct frontmatter, author (mateo), and slug.
  • No broken internal links.
  • Truncate marker (`{/* truncate */}`) shows a reasonable preview on the blog index.

🤖 Generated with Claude Code


Note

Low Risk
New static blog content only; no runtime, auth, or application logic changes.

Overview
Adds a new blog post at blog/agent_shin_triage/index.md (slug agent-shin-triage) that documents Agent Shin, the automated triage bot for external PRs and issues on BerriAI/litellm.

The post explains why triage exists at repo scale, the three outcomes (pass, 24h grace comment, auto-close), and PR vs issue rubrics—including context plus end-to-end QA proof (with an explicit no-mocking rule for pytest-only claims). It also covers recovery via @agent-shin reconsider, @greptileai, and maintainer override after auto-close.

This is documentation-only; it does not change triage automation in the main repo (the PR notes possible follow-up to align strings in .github/scripts/triage_with_llm.py on BerriAI/litellm).

Reviewed by Cursor Bugbot for commit 0108b5b. Bugbot is set up for automated code reviews on this repo. Configure here.

Adds the public-facing explanation linked from every Agent Shin comment.
Covers what the bot checks, the PR + issue rubrics (including the
non-mocked QA-proof requirement), the 24h grace window, and the
@agent-shin reconsider / @greptileai recovery paths.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
@vercel
Copy link
Copy Markdown

vercel Bot commented May 28, 2026

The latest updates on your projects. Learn more about Vercel for GitHub.

Project Deployment Actions Updated (UTC)
litellm Ready Ready Preview, Comment May 28, 2026 3:39am

Request Review

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant