harness-audit

A command that checks an AI coding setup for problems in hooks, links, tool connections, and skill coverage.

In plain words
What is it for?
Use it to audit an aicrew or similar coding-agent installation. It pauses for user decisions at required checkpoints.
Why use it?
It helps reveal missing or misconfigured parts of the agent environment before they cause unreliable work.

Command

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.

agentmods
npx agentmods add commands/aksoftcode/aicrew/harness-audit
Clone the repo
git clone --depth 1 https://github.com/AKSoftCode/aicrew
Per session 17 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,575 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 1 finding. Scan, not verified.
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.

ModelPer sessionOnce invoked
Fable 5 $0.00017 $0.01575
Opus 5 $0.00009 $0.00788
Sonnet 5 $0.00003 $0.00315
Haiku 4.5 $0.00002 $0.00158

Measured yesterday against content hash e1f193f223e8, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

harness-audit scanned grade A with 1 finding 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 yesterday.

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.

Reads agent configuration directorieslowAgent snooping

.claude/, .codex/, .gemini/ hold keys, settings and other credentials a mod has no legitimate need for.

cat ~/.claude/settings.json | python3 -c "import json,sys; s=json.load(sys.stdin); print(json.dumps(s.get('hooks',{}), indent=2))"

Downgraded: this mod is about security review, or the phrase is quoted, so it is likely naming the pattern rather than instructing it.

skills/commands/harness-audit.md · 172 lines

How it starts

The opening of the file, as written. The whole thing — 172 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Health-check your aicrew install — verifies hooks, symlinks, MCP wiring, and Codex skill coverage.

⚠️ INTERACTIVE CHECKPOINTS — MANDATORY RULE

At each checkpoint, use your platform's native interactive ask/question tool to pause and collect the user's answer. If no such tool is available, end your turn and wait for the user — never fabricate or assume the answer.

Known tools by platform (use if available):

Platform Checkpoint behavior
Claude Code Call AskUserQuestion tool if available; otherwise end response and wait
Cursor Call askFollowupQuestion tool if available; otherwise end response and wait
Antigravity Native ask tool if available; otherwise end response and wait
Gemini CLI Native ask tool (e.g. ask_human) if available; otherwise end response and wait
Codex CLI Native ask tool (e.g. ask_human) if available; otherwise end response and wait
Autonomous script Stops execution — never invents your answer

NEVER skip a checkpoint. NEVER fabricate the user's response.

/harness-audit — AI Harness Self-Audit

Audits the Claude Code / AI agent setup itself: skills, agents, hooks, commands, memory, and platform coverage. Finds gaps, stale configs, and reliability issues.

Mode from $ARGUMENTS: quick (5-10 min read-only) or full (complete + recommendations). Default: quick.


Step 1: Skills & Agents inventory

Check ~/Agents/:

ls ~/Agents/commands/
ls ~/Agents/agents/
ls ~/Agents/hooks/

For each file, verify:

  • Has description: field with trigger conditions only (not workflow summary)
  • Has ⚠️ INTERACTIVE CHECKPOINTS table with all 5 platforms
  • Leads with the generic "use your platform's native interactive ask/question tool" instruction before the platform table
  • Platform table lists known native ask tools (AskUserQuestion, askFollowupQuestion, native ask tool e.g. ask_human) as best-effort hints, not just "end response"
  • Agent cross-references use ~/Agents/agents/ path (not ~/.claude/skills/)

Read the full file on GitHub · 172 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.

  1. yesterday First seen · 172 lines · 17 tokens per session scan A e1f193f223e8

Subscribe to this mod's changes

harness-audit is a command published in the GitHub repository AKSoftCode/aicrew (3 stars, last pushed 2mo ago), licensed MIT. It adds 17 tokens to every session and 1,575 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it A with 1 finding (reads agent configuration directories). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.