claim-audit

claim-audit is a skill for Claude Code from 3xachris/3xa-harness. It costs 101 tokens per session (2,564 once invoked), scanned A, original, MIT.

A review tool that finds factual number claims in project records when they have no source or an unverified label.

In plain words
What is it for?
It is for auditing decision logs, work orders, closeout records, and specifications before others rely on them.
Why use it?
It helps later readers tell which figures they can check and which still need evidence, without requiring them to repeat the original investigation.

Skill for Claude Code

Written for Claude Code: PostToolUse hook event. Also seen: mentions Claude Code.

Part of the harness-audit plugin — 2 skills shipped together

Good fit It is for auditing decision logs, work orders, closeout records, and specifications before others rely on them.

Compare 6 skills from other repositories ↓
Install with agentmods
npx agentmods add skills/3xachris/3xa-harness/claim-audit
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 3xachris/3xa-harness --skill claim-audit
Clone the repo
git clone --depth 1 https://github.com/3xachris/3xa-harness

Made for: Claude Code.

Or install harness-audit, the plugin that ships this one along with the rest of its 2 skills.

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.

agentmods badge for claim-audit

README.md
[![agentmods](https://agentmods.dev/badge/skills/3xachris/3xa-harness/claim-audit/github.svg)](https://agentmods.dev/skills/3xachris/3xa-harness/claim-audit)
Your own site
<a href="https://agentmods.dev/skills/3xachris/3xa-harness/claim-audit"><img src="https://agentmods.dev/badge/skills/3xachris/3xa-harness/claim-audit/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.

agentmods 80×15 button for claim-audit

Your own site · 80×15
<a href="https://agentmods.dev/skills/3xachris/3xa-harness/claim-audit"><img src="https://agentmods.dev/badge/skills/3xachris/3xa-harness/claim-audit.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 101 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 2,564 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.

ModelPer sessionOnce invoked
Fable 5.1 $0.00101 $0.02564
Opus 5 $0.00051 $0.01282
Sonnet 5 $0.00020 $0.00513
Haiku 4.5 $0.00010 $0.00256

Measured 10d ago against content hash 20d7bf5290ac, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-10, from the pricing page.

Security

Grade A, and why

claim-audit 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 10d ago.

The scan reads SKILL.md. This mod also ships 1 executable file (claim_audit.py), listed below but not scanned — reading those needs a real analyzer, not pattern matching.

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.

addons/audit/skills/claim-audit/SKILL.md · 101 lines

How it starts

The opening of the file, as written. The whole thing — 101 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Claim Audit

A record is read by someone deciding whether to act without re-deriving it. A number they can trace — to a path, a date, a measurement, a run — costs them one look. A number that arrives bare costs them the whole derivation, which is the work the record existed to save. So the checkable form is:

Every number written into a record either carries a source, or is marked as unverified.

The script claim_audit.py finds the lines that satisfy neither. It is a candidate detector, not a judge: it reads shape, not meaning, so some of what it returns will be fine as written.

How noisy that is, measured rather than asserted — one run over this repository, reproducible with python addons/audit/skills/claim-audit/claim_audit.py --root . --full: 21 Markdown records scanned in about a tenth of a second, 2 lines left for a human, 3 more auto-classified as restatements (§3, The fifth disposition), and both human-judged lines were false positives on review — budget lines declaring a fuse rather than reporting a measurement. So the per-run reading cost here was five lines, and the finding count was zero. That is one small documentation-heavy repo that had already been cleaned against this tool, not a precision figure for yours; run it on your own records before deciding how much attention it deserves, and before binding it to anything (§5). The first run on an older project is the expensive one — that is what the baseline (§4) exists for.

1. What counts as a source

A line passes when it carries any of these, on the same line as the number:

Signal Example Checked how
A path with a line number verify_closeout.py:16 the file must exist
A filename in backticks the limit lives in `config.json` the file must exist
A link, or a WikiLink [[docs/decisions.md#2026-08-01 Picked the cache]] the file must exist
A date 2026-08-07 shape only
A field naming its origin evidence: artifacts/run.log, source: docs/spec.md shape only
A measurement verb measured, benchmarked, reproduced, profiled, counted, timed shape only
A reference id AC-3, #412, PROJ-88, §4 shape only

Read the full file on GitHub · 101 lines

Files

What ships with it

2 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.

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. 10d ago First seen · 101 lines · 101 tokens per session scan A 20d7bf5290ac

Subscribe to this mod's changes

claim-audit is a skill published in the GitHub repository 3xachris/3xa-harness (17 stars, last pushed 10d ago), licensed MIT. It adds 101 tokens to every session and 2,564 once invoked, about $0.0005 per session on Opus 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.