agentleFS
tellmefrankietellmefrankie/ai-investment-skills/pr-review-prep/SKILL.md

pr-review-prep

Skill14 starsChanged 6 months agoMITView on GitHub

Generate a complete pull request description from your git diff. Includes summary, test plan, risk assessment, and reviewer notes — in under 10 seconds. The description above took 8 minutes to write manually. This skill generates it in 8 seconds and catches the risk flags you forget when you're rushing to ship. Reviewers merge faster when they understand the change. Better descriptions = shorter review cycles. Fully free and open source. Part of the AI Investment Skills Bundle — $29…

  • Commits and pushes

1 other project uses a version of this.

Adopted by

Most recently taken up by

The file

Files
SKILL.md

PR Review Prep

Generate a complete pull request description from your git diff. Includes summary, test plan, risk assessment, and reviewer notes — in under 10 seconds.

What it does

  • Reads git diff main...HEAD (or any branch pair)
  • Classifies changes: new feature, bug fix, refactor, dependency update, config change
  • Writes PR title (under 70 chars), summary bullets, test plan checklist
  • Flags risk areas: database migrations, auth changes, public API surface changes
  • Identifies files that need extra reviewer attention
  • Optionally creates the PR via gh pr create

Usage

Prepare my PR description.
Branch: feat/options-flow-lottery-filter vs main
Include: summary, what changed, test plan, risk level, files to review carefully.
Format for GitHub PR body (markdown).

Example Output

## Summary

- Add real/lottery call separation to options flow analyzer
- Lottery threshold: premium < $0.10 OR delta < 0.05
- Adjusted P/C ratio now excludes lottery volume before sentiment classification
- Anomaly detection baseline unchanged; now runs on adjusted ratio

## What changed

**Core logic** (`src/options/analyzer.ts`)
- `splitRealVsLottery()` — new function, filters options by premium + delta thresholds
- `computeAdjustedPC()` — replaces raw P/C with filtered ratio
- Sentiment thresholds unchanged (< 0.45 = bullish, > 1.5 = bearish)

**Types** (`src/types/index.ts`)
- Added `LotteryMetrics` type: `{ lotteryPct, realCallOI, realPutOI }`

**Tests** (`tests/options.test.ts`)
- 4 new test cases: RXRX 84% lottery inversion, KTOS clean signal, edge cases

## Test Plan

- [ ] Run `pnpm test` — all existing tests pass
- [ ] Manual: run analyzer on RXRX — confirm adj P/C ≠ raw P/C when lottery > 50%
- [ ] Manual: run on KTOS — confirm adj P/C close to raw when lottery < 30%
- [ ] Check Polygon API error handling still works (mock API key test)

## Risk Assessment

Risk level: LOW

- No DB schema changes
- No public API surface changes
- Additive only — raw P/C still computed, adj P/C is additional field
- Backward compatible with existing callers

## Files needing careful review

- `src/options/analyzer.ts` — core logic change, review threshold values
- `src/types/index.ts` — new type, ensure downstream consumers updated

Configuration

Prep my PR:
Base branch: main
Head branch: current (auto-detect)
Style: github  # options: github | gitlab | linear | jira
Risk scan: true  # flag auth, migrations, API changes
Create PR: false  # set true to auto-create via gh CLI
Reviewers: @teammate1, @teammate2

Why not just write it manually?

The description above took 8 minutes to write manually. This skill generates it in 8 seconds and catches the risk flags you forget when you're rushing to ship.

Reviewers merge faster when they understand the change. Better descriptions = shorter review cycles.

Requirements

  • Bash tool access (runs git diff, git log)
  • gh CLI installed (optional, for auto-create)

Free

Fully free and open source.

Part of the AI Investment Skills Bundle — $29 one-time. → https://jaehyunpark.gumroad.com/l/tcyahy

Author

Built after submitting a PR with "fixes stuff" as the description and getting it stuck in review for 3 days.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.