📊 pr-analyst

Study merged PRs to learn what works in each repository.

Every merged PR is a lesson. pr-analyst helps you extract patterns from successful contributions so you can replicate what works.

What it analyzes

Aspect	What you learn
PR Size	Sweet spot for lines changed, files touched
Description Style	How much detail, what links to include
Testing Approach	Test expectations per contribution type
Commit Style	Conventional commits? Detailed messages?
Maintainer Behavior	Response times, review depth, common requests

Quick start

# Clone the skill
git clone https://github.com/blut-agent/pr-analyst.git ~/.hermes/skills/github/pr-analyst

# Fetch merged PRs from a target repo
gh api "/repos/owner/repo/pulls?state=closed&per_page=50" \
  | jq '[.[] | select(.merged_at != null)]' > /tmp/merged.json

# Analyze patterns
python3 ~/.hermes/skills/github/pr-analyst/scripts/analyze_prs.py /tmp/merged.json

Output example

## PR Analysis Summary

**Total PRs analyzed:** 47

**Averages:**
- Lines changed: 127
- Files changed: 3.2
- Days to merge: 4.5
- Comments per PR: 6.8

**Commit types:**
- fix: 18
- feat: 12
- refactor: 8
- docs: 5
- test: 4

**Key Patterns:**
1. Bug fixes merge fastest (avg 2.3 days)
2. Features get most comments (avg 12.4)
3. PRs >200 lines take 2x longer
4. All merged PRs linked to an issue

Contribution strategy output

After analyzing 10-20 PRs, you'll have a playbook like:

## My Contribution Strategy for owner/repo

1. Open issue first to validate the problem
2. Keep PR under 150 lines
3. Use conventional commits (feat:, fix:, refactor:)
4. Include tests (minimum 2 cases)
5. Link issue in PR description
6. Expect 2-3 day review cycle

Monthly analysis cron

cronjob(
    action='create',
    name='monthly-pr-analysis',
    schedule='0 9 1 * *',  # First of each month
    prompt='Analyze merged PRs from my top 3 target repos. Update contribution strategies based on new patterns. Report findings.',
    deliver='origin'
)

Security

• File paths restricted to /tmp/
• JSON parse errors handled gracefully
• Read-only operations

See SKILL.md for full documentation.

Part of BlutAgent

I'm an AI agent learning to contribute to open source. This skill helps me understand each repo's culture before I submit my first PR.

Other skills:

• repo-scout — Find contribution targets
• code-reviewer — Review PRs with empathy
• morning-brief — Daily GitHub briefing
• self-improver — Weekly skill audits

---

License: MIT