ponytail-audit
Quality audit of a whole repo: bugs, security holes, what breaks under real load, risky code without tests, slow paths, and what to delete, merge or split. One-shot report, changes nothing. Run only when the user invokes /ponytail-audit by name, or a skill points at this file.
- License 许可
- MIT
- 最近更新
- 2026-11-10
- Stars
- ★ 0
- Forks
- 0
01 前言 Front Matter
- name
- ponytail-audit
- metadata
- { "fork": { "change": "Copied from ponytail without its hooks, so it never runs by itself. User-invoked, and the trigger phrases are out of the description." } }
- description
- Quality audit of a whole repo: bugs, security holes, what breaks under real load, risky code without tests, slow paths, and what to delete, merge or split. One-shot report, changes nothing. Run only when the user invokes /ponytail-audit by name, or a skill points at this file.
- disable-model-invocation
- true
02 资源文件 2
- 参考文档agents/openai.yaml168 B
- 许可证LICENSE1.0 KB
03 SKILL.md 正文
Audit the whole repo like the senior developer who just inherited it and will be paged when it breaks. Order of importance: correct, safe, holds under load, tested, fast, lean. Lean still matters: every extra line must be read, tested and fixed later. This is a report the user asked for, so give it in full.
1. Map first
- Audit what the user names: a folder, a package, or the whole repo. Nothing named: the whole repo.
- Read the README, the deploy and build config, the dependency list, the entry points (main, routes, handlers, jobs, CLI commands) and the tests.
- Find the expected load: one person running a script, or many users and processes at once. Judge scale against that, and say which load you assumed.
- Trace the main flows end to end: where data comes in, what is stored, what goes out. Read those paths fully: input from users, money, auth, data writes, background jobs, anything shared between processes.
- Big repo: go deep where a mistake costs the most, not file by file. Say which parts you did not read.
2. Look for
- Bug: wrong result, crash, missed edge case (empty, zero, last item, rounding, time zones), callers that disagree with what a function returns, the same rule applied differently in two places.
- Risk: security holes (injection, weak randomness, secrets in code, missing checks on input from users), data loss (errors swallowed, writes in the wrong order, no transaction).
- Scale: fine for one user, wrong for many: check-then-write races, the same work done by every process, memory or lists that only grow, a query per item, O(n^2) on big input, per-process state that must be shared.
- Missing test: risky logic (a branch, a parser, money, security, data writes) with no test that fails when it breaks. One good test, not coverage.
- Speed: big slowdowns are problems. Small wins (work repeated in a hot loop) are suggestions; some software counts every millisecond.
- Lean: code that should not exist or should be smaller.
- delete: dead code, unused options, flags and config, speculative features
- reuse: two helpers doing the same thing (keep one, name the path)
- stdlib / native: the standard library or platform already does it; a dependency doing what a few lines or the platform can do
- yagni: interface with one implementation, factory with one product, wrapper that only passes calls through
- merge: near-copies that must change together
- split: one function or class doing several unrelated jobs, so it is hard to read or test. Split by job, never by line count, and never into helpers that exist only to make a function shorter.
3. Check before you report
- Every finding needs a concrete case: "this input or situation leads to this wrong result". No case, no finding.
- Before calling code unused, grep the whole tree for it, including tests, fixtures, config, and string or dynamic references.
- A shortcut marked with a
shortcut:(or olderponytail:) comment that names its limit is a decision, not a finding, unless the expected load already crosses it. - Propose the smallest fix that works. Prefer fixes that delete code. Never add layers, frameworks or config the problem does not need.
- No style taste, no "consider", no vague worries.
4. Output
Very simple English: short sentences, everyday words. Explain a technical term the first time you use it. The reader may never have seen this code.
Start with What this repo does: in two or three sentences, and the load
you assumed.
Then the findings in three groups, most important first, skip empty groups:
- Must fix: bug, security, data loss, breaks at the expected load.
- Should fix: risky code without a test, real slowness, duplication, a function that mixes jobs, code that should not exist.
- Nice to have: small speed-ups, shorter forms.
Number findings across all groups, so the user can say "fix 2 and 5". At most 20 findings; if you left smaller ones out, say how many. Every finding has all four parts, each one or two short sentences:
- Orders land on the wrong day (
billing/close_day.py:L40-52)- What this is: At midnight this job closes the day and bills all orders of that day.
- Problem: It takes "today" from the server clock, which runs in UTC. An order placed at 00:30 in Berlin is billed on the day before.
- Fix: Compute the day once in the shop's time zone:
datetime.now(ZoneInfo("Europe/Berlin")).date(). One line, nothing else changes. - If we skip it: Late orders show the wrong date, and accounting fixes them by hand.
End with:
Verdict:one line: healthy, or what to fix first.Lean: -<N> lines, -<M> dependencies possible.when lean findings exist.Not checked:the parts you did not read or could not run.
Nothing found: What this repo does:, then Healthy. Nothing to fix. and
one line on what you checked.
One-shot report, changes no code.
04 同仓库更多 Skills
improve
暂无简介
create-verification-skill
Generate a project-local verification skill that drives your app the way a user does — any language, framework, or platform. Use for /create-verification-skill, "make a control skill for this repo", or when a project has no scripted way to prove UI/CLI/service behavior.
tdd
Test-driven development. Use when the user wants to build features or fix bugs test-first, mentions "red-green-refactor", or wants integration tests.
maintain-verification-skill
Periodic pass that keeps a project's verification skill and feature map honest: parallel source readers per feature, one live session driving every feature, at most one PR of proven corrections. Use for /maintain-verification-skill or "audit the verify skill".
verdict
The Reviewer's Verdict on a PR: re-run the repo's verification skill at the PR's head, hold the PR's Evidence against it, and record pass or fail with nico-agent.
research
Investigate a question against high-trust primary sources and capture the findings as a Markdown file in the repo. Use when the user wants a topic researched, docs or API facts gathered, or reading legwork delegated to a background agent.