Pr generator

AI Engineering Operating System

Install

npx -y skills add willianbs/skills --skill pr-generator

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

2 things to look at

20 days oldThe repository was created 20 days ago. New is not bad, but a brand new repository carrying a familiar-sounding name is the shape a typosquat arrives in, and there has been no time for anyone else to find a problem with it.
0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

Writes accurate PR descriptions from git diff and commits only—never invented intent or test results. Use when opening or updating a pull request. Emits PR_BODY. Prefer the repo’s PR template when present.

SKILL.md

2.6 KB, 561 tokens by cl100k_base, as published. Nobody here has run it

Purpose

Produce a reviewer-ready PR description that is archaeologically honest.

When to Use / When NOT to Use

Use when: creating/updating a PR; after IMPL_REPORT / QG_REPORT.

Do not use when: no diff yet; user wants marketing copy unrelated to changes.

Preconditions

Must read git diff / commits (and PR template if any). Refuse to invent a body without the change set.

Inputs / Outputs

Inputs: diff, commits, IMPL_REPORT, QG_REPORT/REVIEW (optional), ticket ID if provided.

Outputs: PR_BODY (markdown)

Upstream / Downstream

Upstream: feature-implementer, quality-gate, code-reviewer.

Downstream: human/agent opens PR (gh pr create).

Core Principles

Diff and commits are the source of truth.
Unknown beats fabricated “Why.”
Testing section lists only commands actually run.
Prefer .github/PULL_REQUEST_TEMPLATE.md (or org template) structure.
Call out risks and rollout notes when evidenced.
No emoji stuffing or hype.
Never claim QG Pass without QG_REPORT.

Process

Load template if present; else use default sections below.
Summarize what changed from diff (paths + behavior).
Why — only from commits/ticket/user; else Unknown — please fill.
Test plan — from IMPL_REPORT commands_run only; mark untested explicitly.
Risks / rollback — from PLAN or obvious migration/flag evidence.
Include QG/Review decision if artifacts exist.
Reviewer focus — sharp bullets, not essays.

Evidence Requirements

Every factual claim cites path or commit subject. Omit unsupported sections rather than inventing.

Stop Conditions / Failure Modes

Condition	Action
No diff	Refuse
User asks to hide risks	Refuse; include risks or stop

Severity + Confidence

N/A. Accuracy over completeness.

Output Contract

Default markdown:

## Summary
...

## Why
...

## Changes
- `path`: ...

## Test plan
- [ ] `command` — result (from IMPL_REPORT) / NOT RUN

## Risks & rollback
...

## Quality
QG/Review: ... or Unknown

## Reviewer focus
...

Handoffs

None required. Optionally note remaining Block items from QG.

Never

Never invent business intent, metrics, or test results.
Never omit breaking changes visible in the diff.
Never include secrets from the diff; redact and warn.

What ships with it: 1 file

672 B alongside SKILL.md

README.md672 B

Gives 0 of the 12 instructions most pr commit review skills give in 561 tokens

Counted across 888 of the 1,342 authors here whose files we hold, read 2026-08-07

Use conventional commits formatin 127 of 888, across 115 files
Keep subject line under 72 charactersin 62 of 888, across 48 files
Delete branches after mergein 51 of 888, across 38 files
Use imperative mood in subject linein 51 of 888, across 42 files
Use imperative mood in commit messagesin 44 of 888
Verify directory is ignored before creating worktreein 43 of 888, across 12 files
Generate a conventional commit messagein 43 of 888
Add unignored worktree directories to gitignorein 42 of 888, across 10 files
Make atomic commitsin 39 of 888, across 27 files
Run tests before committingin 36 of 888, across 25 files
Verify clean test baselinein 35 of 888, across 9 files
Split unrelated changes into separate commitsin 35 of 888, across 30 files

Said here and by no other author read

summarize what changed from the diff
derive the rationale only from commits or tickets
list only commands actually run for testing
cite path or commit subject for every factual claim
include risks and rollback notes when evidenced
include quality gate or review decisions when artifacts exist

Grouped from the skills themselves: near-identical wordings counted once, and counted by distinct author, so one author publishing three of these counts once. Length counted with cl100k_base; the agent that loads this file may tokenize it differently.