What is audit-harness?

A guide for checking the harness files and automated checks of tunacode-cli. These checks include commit hooks, push hooks, architecture rules, and continuous-integration workflows.

Use it to audit HARNESS.md, verify hooks and CI workflows, check architecture gates, and explain what actually enforces the project's rules.

It treats skipped, mismatched, or failing checks as important findings and verifies each required check directly, so a summary or wrapper command cannot hide a problem.

audit-harness — Skill by alchemiststudiosDOTai/tunacode

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

Any agent

npx skills add alchemiststudiosDOTai/tunacode --skill audit-harness

Clone the repo

git clone --depth 1 https://github.com/alchemiststudiosDOTai/tunacode

Made for: Claude Code, Codex.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

README.md

[![agentmods](https://agentmods.dev/badge/skills/alchemiststudiosdotai/tunacode/audit-harness/github.svg)](https://agentmods.dev/skills/alchemiststudiosdotai/tunacode/audit-harness)

Your own site

<a href="https://agentmods.dev/skills/alchemiststudiosdotai/tunacode/audit-harness"><img src="https://agentmods.dev/badge/skills/alchemiststudiosdotai/tunacode/audit-harness/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

Your own site · 80×15

<a href="https://agentmods.dev/skills/alchemiststudiosdotai/tunacode/audit-harness"><img src="https://agentmods.dev/badge/skills/alchemiststudiosdotai/tunacode/audit-harness.svg" alt="Reviewed on agentmods" width="80" height="20"></a>

Per session 69 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.

When invoked 746 The whole file, excluding the scripts and references it only reads on demand.

Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe. Third-party audits

NVIDIA SkillSpector pass 7 Sept 2026

How audits are shown

Origin original No closer match found in the catalogue.

Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

Model	Per session	Once invoked
Fable 5.1	$0.00069	$0.00746
Opus 5.5	$0.00028	$0.00298
Sonnet 5.5	$0.00014	$0.00149
Haiku 4.5	$0.00007	$0.00075

Measured 1mo ago against content hash be146927dfad, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

audit-harness scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 1mo ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

docs/skills/audit-harness/SKILL.md · 102 lines

How it starts

The opening of the file, as written. The whole thing — 102 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Audit Harness

HARNESS.md is mission-critical. Audit it with zero tolerance.

Trigger

Use this skill when the user asks to:

audit HARNESS.md
verify harness accuracy
run pre-commit or pre-push hooks manually
confirm architecture or CI gates
explain what actually enforces the harness

Hard Rules

Treat any mismatch, omission, skipped gate, auto-fix, or failing check as a critical failure.
Never describe results as "mostly passing", "just one failure", or equivalent minimization.
State the exact failure first.
Explain the finding before changing code or docs.
Do not proceed with a fix until the user tells you to proceed.
During harness audits, never use make check, scripts/run_gates.py, or any loop/script wrapper as the primary audit path.
Run checks manually, one by one, in the same order they appear in the source-of-truth config.
If a hook modifies files, report the exact files immediately.
Do not revert hook changes unless the user explicitly asks.

Source Of Truth Order

Read these first:

HARNESS.md
.pre-commit-config.yaml
Makefile
tests/test_dependency_layers.py
scripts/grimp_layers_report.py
.github/workflows/*.yml
docs/git/practices.md
AGENTS.md

Manual Audit Procedure

Pre-commit

Enumerate the active pre-commit hooks from .pre-commit-config.yaml.
Start at the top.
Run each hook manually:

uv run pre-commit run <hook-id> --all-files

After each hook: state Passed, Failed, Skipped, or Modified files.
If a hook fails, stop and explain why before proposing a fix.

Pre-push

Enumerate the active pre-push hooks from .pre-commit-config.yaml.
Run each one manually, one by one:

uv run pre-commit run <hook-id> --hook-stage pre-push --all-files

Treat any failure as critical.

Architecture

tests/test_dependency_layers.py is the source of truth for grimp enforcement.
scripts/grimp_layers_report.py is report generation only.
scripts/run_gates.py is supplemental only and not canonical.

Read the full file on GitHub · 102 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

1mo ago First seen · 102 lines · 69 tokens per session scan A be146927dfad

Subscribe to this mod's changes

audit-harness is a skill published in the GitHub repository alchemiststudiosDOTai/tunacode (125 stars, last pushed 15d ago), licensed MIT. It adds 69 tokens to every session and 746 once invoked, about $0.0003 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.

CI and CD DevOps and cloud LLM Testing agent ai anthropic cli developer-tools

knowledge-engineering-quality-and-delivery

A project-wide quality and delivery workflow covering tests, release checks, builds, publishing, and bilingual documentation for Blade Code.

echoVic/blade-code · 132 tokens

evaluating-llms-harness

Evaluates LLMs across 60+ academic benchmarks (MMLU, HumanEval, GSM8K, TruthfulQA, HellaSwag). Use when benchmarking model quality, comparing models, reporting academic results, or tracking training progress. Industry standard used by EleutherAI, HuggingFace, and major labs. Supports HuggingFace, vLLM, APIs.

StarryCod/cogitum · 29 tokens

dogfood

Exploratory QA of web apps: find bugs, evidence, reports.

StarryCod/cogitum · 18 tokens

test-driven-development

TDD: enforce RED-GREEN-REFACTOR, tests before code.

StarryCod/cogitum · 20 tokens

baby-sit

Monitor a GitHub pull request until CI is green, diagnose failures, and rerun only evidence-backed flaky GitHub Actions jobs.

langchain-ai/open-swe · 30 tokens

reproduce

Reproduces a paper's result, a claim, an artifact or a previous run with the target and success criterion frozen first, the canonical code path run before any substitute, exact inputs, environment, seeds and commands captured, and a verdict from a fixed set, reproduced, reproduced within tolerance, partially…

synthetic-sciences/openscience · 116 tokens