verify SKILL.md skillA

verify from ismaildasci/claude-referee is an agent skill: a SKILL.md folder an AI agent loads when a task calls for it.

Install it, pinned
mdr add ismaildasci/claude-referee/verify@git:20261007.ea34306

Pins git:20261007.ea34306 in mdr.lock. Pin the fingerprint instead.

Audit grade
A, 17 of 17 checks passed
Kind
SKILL.md skill
Repository
ismaildasci/claude-referee
Path
docs/recipes/verify/SKILL.md
Versions recorded
6
Latest version
git:20261007.ea34306, committed Oct 7, 2026
First version recorded
Oct 1, 2026
Size
1,366 bytes
Repository stars
8
License
MIT

Check that the work is really done before committing. Runs the project's test command and pipes its output to claude-referee done, so the verdict comes from the evidence and not from a summary.

Collected from public GitHub. Every version is saved exactly as it was and never changes. Install it at an exact version with the mdr tool, and every file is checked before it reaches your agent. The main file is checked against the SHA-256 fingerprint recorded here, and the others against the fingerprints GitHub keeps for its source commit. The automatic checks below grade the latest version, and the same checks always give the same file the same grade.

What the file says

# verify

Run this right before committing code changes.

1. Run the project's check command and pipe all of its output, including stderr and the exit code, to the referee. Replace `npm test` with the project's own command (for example `cargo nextest run`, `pytest`, `go test ./...`). The exit code line keeps the real status (a failing run is `missing`); for a runner the referee does not recognise, only the exit code is read, so a passing run often comes back `unsure` or `missing`, and a `met` proves only the exit status:

```bash
{ npm test; echo "exit code: $?"; } 2>&1 | npx claude-referee done --criteria "all tests pass" --evidence - --fail-on missing,unsure
```

2. Read only the one-line JSON it prints.
   - `met`: commit.
   - `missing` or `unsure`: do not commit. Follow `next_step`, fix the cause, and run step 1 again.
   - No JSON line (npx could not run the package, a missing key, a crash): there is no verdict. Do not retry in a loop and do not claim success. Tell the user the check could not run and why.
3. Do not paraphrase the test output to claim success. The verdict line is the evidence.

Read the whole file at its exact version.

How to install

The exact contents, by fingerprint
mdr add ismaildasci/claude-referee/verify@sha256:a8a375377015c96f

The line at the top of this page pins the label git:20261007.ea34306, which is what most people want. This one pins the sha256 fingerprint of the contents instead, so it survives a label being moved. Either way the exact version is written to mdr.lock, and mdr install downloads and checks those contents again, so it installs them exactly or stops with an error.

Badge

mdr badge

[![mdr](https://markdownregistry.com/badge/art_ww5sdfodccuqmnkn.svg)](https://markdownregistry.com/a/art_ww5sdfodccuqmnkn)

2 badge image fetches in 30 days. What the badge shows.

Versions

The registry has recorded 6 versions of this file, between Oct 1, 2026 and Oct 7, 2026, about one every day.

versioncommittedcommitsizeaudit
git:20261007.ea34306 latest2026-10-07 ea34306 1,366 BA view · diff
git:20261006.efc154b2026-10-06 efc154b 1,299 BA view · diff
git:20261005.baa8ccf2026-10-05 baa8ccf 1,496 BA view · diff
git:20261005.2e65abc2026-10-05 2e65abc 1,491 BA view · diff
git:20261001.235ebd82026-10-01 235ebd8 1,434 BA view · diff
git:20261001.d3bb2c32026-10-01 d3bb2c3 1,424 BA view

Audit of the latest version

A  17 of 17 checks passed. Fixed automatic checks with no AI involved, so the same checks give the same answer every run.
  • pass: Frontmatter block present
  • pass: Frontmatter declares a name
  • pass: Frontmatter declares a description
  • pass: Size between 200 bytes and 200 KB (1366 bytes)
  • pass: No zero-width or bidi control characters
  • pass: No instruction hidden inside an HTML comment
  • pass: No link to an exfiltration or paste host
  • pass: No credential-shaped string
  • pass: No instruction to send local credentials anywhere
  • pass: No text hidden with inline styles
  • pass: No prompt-injection phrasing
  • pass: No curl or wget piped into a shell
  • pass: No recursive delete of root, home or parent
  • pass: No instruction to read or print local credentials
  • pass: No base64 blob over 200 characters
  • pass: No link to a raw IP address
  • pass: No script tag

Source

GitHub

ismaildasci/claude-referee · 8 stars · license MIT · pushed 2026-10-21 · branch main

API

GET https://markdownregistry.com/api/v1/artifacts/art_ww5sdfodccuqmnkn
GET https://markdownregistry.com/api/v1/resolve?ref=ismaildasci/claude-referee/verify
GET https://markdownregistry.com/api/v1/blob/a8a375377015c96f2005f2f643191b430cb1ee5ad7fbbdb1adaf7ed71210aedf

Your AI assistant does the legwork. You hear about the deals worth your word. Hand yours the standing instructions at modelranch.com and it joins the network that reads files like this one.

Guides: the SKILL.md format and its required fields · what to check before you install a skill · how to pin a skill to an exact version · where to find agent skills · keeping a team on the same skills · which agent file does what.

More from ismaildasci/claude-referee

jev SKILL.md skill
ismaildasci/claude-referee · plugins/claude-referee/skills/jev/SKILL.md · Hand small, checkable judgements to TypeSafe Jev through the claude-referee CLI instead of judging them yourself. done…
git:20261015.71bc158 · audit A · 8 stars

Every file in ismaildasci/claude-referee

Other files named verify

verify SKILL.md skill
yeachan-heo/oh-my-claudecode · skills/verify/SKILL.md · Verify that a change really works before you claim completion
git:20260401.ff9223c · audit A · 40,148 stars
verify SKILL.md skill
zenbu-labs/terminal-browser · .claude/skills/verify/SKILL.md · Drive an engine app headlessly in a pty, record a video of the whole verification, and open a summary page (video +…
git:20260721.c3339dc · audit A · 4,102 stars
verify SKILL.md skill
alsk1992/cloddsbot · src/skills/bundled/verify/SKILL.md · Verify agent identity using ERC-8004 on-chain registry
git:20260130.5926fbf · audit A · 2,990 stars
verify SKILL.md skill
zereight/gitlab-mcp · .github/skills/verify/SKILL.md · Verify that a change really works before claiming completion. Activate when user says: verify this, check if it works…
git:20260412.9d9ce51 · audit A · 2,059 stars
verify SKILL.md skill
ethanyoq/ai-novel-writer · .claude/skills/verify/SKILL.md · 在真实 Electron 中驱动 AI 小说作家,按场景截图、记录启动耗时并读取只读诊断。前端重设计 RD-0 至 RD-10 每个 PR 的实机验证都走这里。用于「实机验证」「截图对比」「跑一下界面场景」。
git:20261021.a8f60fa · audit A · 1,705 stars
verify SKILL.md skill
codeaholicguy/ai-devkit · skills/verify/SKILL.md · AI DevKit · Enforce evidence-based completion claims — require fresh command output before reporting success. Use when…
git:20260703.7628b9e · audit A · 1,648 stars
verify SKILL.md skill
sethgammon/citadel · skills/verify/SKILL.md · Self-test the Citadel hook pipeline from within a live session. Exercises real tool calls (Write, Edit, Bash, Read) and…
git:20260919.f58febd · audit A · 925 stars
verify SKILL.md skill
oliver-kriska/claude-elixir-phoenix · plugins/rewrite/skills/verify/SKILL.md · Verify a rewrite slice against its obligations: preflight, capture fresh test evidence, audit translationese and…
git:20261016.88cbbaa · audit A · 569 stars

Browse by kind, by grade A, or by owner.