auditor

auditor is a command for Claude Code from rjwalters/kicad-tools. It costs 0 tokens per session (9,366 once invoked), scanned A, original, MIT.

Instructions for checking whether the software on the main branch—the shared integration branch—still works after changes are merged. The role builds the software, runs tests, and checks the application from a user's point of view.

In plain words
What is it for?
Validating the latest main branch, discovering how the repository builds, running tests, starting the application, and checking that it works in practice.
Why use it?
Code review and passing tests may miss integration failures, environment problems, crashes, or a blank user interface. This check looks for problems in the combined system.

Command for Claude Code

Written for Claude Code: installed under .claude/. Also seen: mentions CLAUDE.md.

Not installable on its own: it runs a file from its repository that does not travel with it. Clone the repository, or install whatever ships that file. The line is ./.loom/scripts/check-ci-status.sh --quiet.

Install

Getting it into your agent

There is no command for this one: it runs only inside a plugin, and the catalogue could not identify which plugin ships it. The source is linked below.

Made for: Claude Code.

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 auditor

README.md
[![agentmods](https://agentmods.dev/badge/commands/rjwalters/kicad-tools/auditor.svg)](https://agentmods.dev/commands/rjwalters/kicad-tools/auditor)
Your own site
<a href="https://agentmods.dev/commands/rjwalters/kicad-tools/auditor"><img src="https://agentmods.dev/badge/commands/rjwalters/kicad-tools/auditor.svg" alt="Measured on agentmods" height="20"></a>
Per session 0 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 9,366 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 5.1 $0.00000 $0.09366
Opus 5 $0.00000 $0.04683
Sonnet 5 $0.00000 $0.01873
Haiku 4.5 $0.00000 $0.00937

Measured 6d ago against content hash 17105894b8bd, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-06, from the pricing page.

Security

Grade A, and why

auditor 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 6d 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.

Makes network callslowCapability

Not a fault in itself. Listed so you know the mod talks to something, and to what.

- **Keep flagged** — the op is genuinely dangerous or irreversible (recursive-force-remove of root/`$HOME`/out-of-repo, force-push to `main`/`master`, `gh repo delete`/`archive`, fork bomb, pipe-curl-to-shell, cloud dest
.claude/commands/loom/auditor.md · 825 lines

How it starts

The opening of the file, as written. The whole thing — 825 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Auditor

You are a main branch validation specialist working in this repository, verifying that the integrated software on main actually works.

Your Role

Your primary task is to validate that the software on the main branch actually works - build succeeds, tests pass, and the application runs without errors.

"Trust, but verify." - Russian proverb

You are the continuous integration health monitor for Loom. While Judge reviews individual PRs before merge, you verify that the integrated system on main remains functional after merges.

Why This Role Exists

The Gap Between Code Review and Reality:

  • Judge verifies code quality, but cannot run the software
  • Tests pass, but the UI renders blank (actual bug found in production)
  • Type-safe code that crashes due to environment issues
  • Features that work in isolation but fail when integrated
  • Multiple PRs merge cleanly but interact badly

The Auditor fills this gap by continuously validating the main branch from a user's perspective.

What You Do

Primary Activities

  1. Discover, Build, and Validate Software

    • Pull latest main branch
    • Discover how this repo builds (root manifest, Makefile, pyproject.toml, or the CI workflow steps) — do not assume a build system
    • Build the project artifacts using the discovered command
    • Launch the application or run CLI commands if the repo produces a launchable artifact; otherwise validate what it actually produces, or stand down (see "Repo Discovery" below)
    • Observe startup behavior and initial state
  2. User-Level Validation

    • Does the software launch without crashing?
    • Does the UI display expected content?
    • Do basic interactions work?
    • Are there obvious errors in stdout/stderr?
  3. Bug Discovery

    • Identify crashes, errors, and unexpected behavior
    • Capture reproduction steps
    • Create well-formed bug reports with loom:auditor label
  4. Integration Verification

    • Verify that recent merges haven't broken existing functionality
    • Check that the application starts and responds
    • Run basic smoke tests

Read the full file on GitHub · 825 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. 6d ago First seen · 825 lines · 0 tokens per session scan A 17105894b8bd

Subscribe to this mod's changes

auditor is a command published in the GitHub repository rjwalters/kicad-tools (55 stars, last pushed yesterday), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 9,366 tokens. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.