audit-harness SKILL.md skillA
audit-harness from guilhermebkel/harness-ledger is an agent skill: a SKILL.md folder an AI agent loads when a task calls for it.
mdr add guilhermebkel/harness-ledger/audit-harness@git:20261024.5a8e6acPins git:20261024.5a8e6ac in mdr.lock. Pin the fingerprint instead.
- Audit grade
- A, 17 of 17 checks passed
- Kind
- SKILL.md skill
- Repository
- guilhermebkel/harness-ledger
- Path
skills/audit-harness/SKILL.md- Versions recorded
- 5
- Latest version
- git:20261024.5a8e6ac, committed Oct 24, 2026
- First version recorded
- Oct 4, 2026
- Size
- 15,441 bytes
- Repository stars
- 5
- License
- MIT
Analyzes this project's coding-agent harness (instruction files, skills, subagents, commands, hooks, MCP servers, plugins) against the session transcripts already saved on this machine, and suggests evidence-based changes. Use when the user asks to improve, audit or clean up their harness, asks why the agent keeps making the same mistake, asks whether a skill, subagent or instruction is worth it, or asks whether a harness change helped.
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
# audit-harness You turn deterministic signals from the user's own sessions into a short list of classified, evidence-backed suggestions for their harness. The numbers come from a local script; your job is to judge, classify and write the change. Never invent a number, a session or a step. The request: $ARGUMENTS (when this line shows no text or the literal variable, use what the user just asked) ## The model The script counts; you judge. Grouping signals by cause, telling real friction from ordinary work and choosing the right fix is the judgment this skill exists for. Don't comment on which model you are running on. ## The script All data comes from the bundled script, `scripts/harness-ledger.mjs` inside this skill's folder (the folder that holds this `SKILL.md`; take the path from where the skill was loaded). Run it with Bash from the project root: ```bash node "<this skill's folder>/scripts/harness-ledger.mjs" <command> --provider <type> [options] ``` `<type>` is the agent you are running in: `claude-code`, `codex` or `cursor`. Pass it on every command, so the script reads that agent's sessions. …
Read the whole file at its exact version.
How to install
mdr add guilhermebkel/harness-ledger/audit-harness@sha256:f1398e2c16960284The line at the top of this page pins the label git:20261024.5a8e6ac, 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_7uijmgvay5dfeprn)
1 badge image fetches in 30 days. What the badge shows.
Versions
The registry has recorded 5 versions of this file, between Oct 4, 2026 and Oct 24, 2026, about one every 5 days.
| version | committed | commit | size | audit | |
|---|---|---|---|---|---|
| git:20261024.5a8e6ac latest | 2026-10-24 | 5a8e6ac | 15,441 B | A | view · diff |
| git:20261021.2fd3595 | 2026-10-21 | 2fd3595 | 15,453 B | A | view · diff |
| git:20261004.c2982c4 | 2026-10-04 | c2982c4 | 14,747 B | A | view · diff |
| git:20261004.7ae17cf | 2026-10-04 | 7ae17cf | 13,825 B | A | view · diff |
| git:20261004.656d4b9 | 2026-10-04 | 656d4b9 | 13,813 B | A | view |
Audit of the latest version
- pass: Frontmatter block present
- pass: Frontmatter declares a name
- pass: Frontmatter declares a description
- pass: Size between 200 bytes and 200 KB (15441 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
guilhermebkel/harness-ledger · 5 stars · license MIT · pushed 2026-10-25 · branch master
API
GET https://markdownregistry.com/api/v1/artifacts/art_7uijmgvay5dfeprn GET https://markdownregistry.com/api/v1/resolve?ref=guilhermebkel/harness-ledger/audit-harness GET https://markdownregistry.com/api/v1/blob/f1398e2c16960284a83f79754e46f6caf18251a1aff1a579ee986e2947798a28
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.