judge-audit SKILL.md skillB
judge-audit from ianalloway/openclaw-skills is an agent skill: a SKILL.md folder an AI agent loads when a task calls for it.
mdr add ianalloway/openclaw-skills/judge-audit@git:20261031.825de9ePins git:20261031.825de9e in mdr.lock. Pin the fingerprint instead.
- Audit grade
- B, 16 of 17 checks passed
- Kind
- SKILL.md skill
- Repository
- ianalloway/openclaw-skills
- Path
judge-audit/SKILL.md- Versions recorded
- 2
- Latest version
- git:20261031.825de9e, committed Oct 31, 2026
- First version recorded
- Sep 14, 2026
- Size
- 7,333 bytes
- Repository stars
- 4
- License
- MIT
Audit an LLM-as-judge with juryrig — position bias, verbosity bias, prompt injection, consistency, and calibration via CLI or HttpJudge.
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
# Judge Audit (juryrig) Audit an **LLM-as-judge** before you trust its scores. Uses the zero-dependency [`juryrig`](https://github.com/ianalloway/juryrig) package to measure position bias, verbosity bias, prompt-injection susceptibility, self-consistency, panels, and calibration — including against local OpenAI-compatible endpoints via `HttpJudge`. ## When to Use Use this skill when: - You are about to ship or rely on an **LLM-as-judge** eval loop - A judge’s scores look too good (or too noisy) and need a bias check - You want a **CI gate** (`exit 1` when flagged) before promoting a judge model - You need to point audits at **Ollama / vLLM / LM Studio / OpenRouter** (or any OpenAI chat-completions URL) without vendor-specific SDKs Skip it for one-off manual grading with a human in the loop, or when you only need a single score and are not treating the model as a trusted judge. ## Install juryrig ```bash pip install juryrig # requires Python 3.10+ juryrig --help ``` Package: [pypi.org/project/juryrig](https://pypi.org/project/juryrig/). Source: [github.com/ianalloway/juryrig](https://github.com/ianalloway/juryrig). ## Prepare `cases.json` …
Read the whole file at its exact version.
How to install
mdr add ianalloway/openclaw-skills/judge-audit@sha256:b98456c9c4d4ddecThe line at the top of this page pins the label git:20261031.825de9e, 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.
[](https://markdownregistry.com/a/art_i3adk6m3h4yvu56v)
3 badge image fetches in 30 days. What the badge shows.
Versions
The registry has recorded 2 versions of this file, between Sep 14, 2026 and Oct 31, 2026, about one every 47 days.
| version | committed | commit | size | audit | |
|---|---|---|---|---|---|
| git:20261031.825de9e latest | 2026-10-31 | 825de9e | 7,333 B | B | view · diff |
| git:20260914.2d79338 | 2026-09-14 | 2d79338 | 7,074 B | B | view |
Audit of the latest version
- fail: No link to a raw IP address (matched: http://127.0.0.1:11434)
- pass: Frontmatter block present
- pass: Frontmatter declares a name
- pass: Frontmatter declares a description
- pass: Size between 200 bytes and 200 KB (7333 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 script tag
Starting a new one? Use a template for SKILL.md on mdpantry, which passes every one of these checks and says what to change in it.
Source
ianalloway/openclaw-skills · 4 stars · license MIT · pushed 2026-11-05 · branch main
API
GET https://markdownregistry.com/api/v1/artifacts/art_i3adk6m3h4yvu56v GET https://markdownregistry.com/api/v1/resolve?ref=ianalloway/openclaw-skills/judge-audit GET https://markdownregistry.com/api/v1/blob/b98456c9c4d4ddec762d4399b578ef5141e638ccfac6ccc9a1de2c8a4755e241
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.