What is cc-audit CLAUDE.md?

Developer notes for cc-audit, a local tool that reads Claude Code transcripts and reports spending attribution, model usage choices, and AI-work habits. Its normal analysis runs on the user's machine, while sending data elsewhere requires explicit consent.

Use them when working on transcript parsing, cost or model attribution, reports, consent-controlled network actions, CLI builds, packaging, linting, type checks, or tests.

They explain the tool's privacy boundary and help preserve it while changing parsing, reporting, commands, or build scripts.

How pa-arth/cc-audit configures Claude Code and Codex

Reuse

Borrowing it

Nothing to install: this file belongs to pa-arth/cc-audit. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.

Copy the file

curl -O https://raw.githubusercontent.com/pa-arth/cc-audit/main/CLAUDE.md

Clone the repo

git clone --depth 1 https://github.com/pa-arth/cc-audit

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/instructions/pa-arth/cc-audit/claude-md.svg)](https://agentmods.dev/instructions/pa-arth/cc-audit/claude-md)

Your own site

<a href="https://agentmods.dev/instructions/pa-arth/cc-audit/claude-md"><img src="https://agentmods.dev/badge/instructions/pa-arth/cc-audit/claude-md.svg" alt="Measured on agentmods" height="20"></a>

Per session 2,869 This file is loaded in full into every session.

When invoked 2,869 The same file — it is already loaded in full.

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

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	$0.02869	$0.02869
Opus 5	$0.01435	$0.01435
Sonnet 5	$0.00574	$0.00574
Haiku 4.5	$0.00287	$0.00287

Measured 7d ago against content hash be9637eba7d6, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-06, from the pricing page.

Security

Grade A, and why

cc-audit CLAUDE.md 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 7d 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.

CLAUDE.md · 168 lines

How it starts

The opening of the file, as written. The whole thing — 168 lines — stays where its author put it; the contents beside it link to each section on GitHub.

cc-audit

@promptster/cc-audit — point it at your local Claude Code transcripts and see where the money and the bad habits are. Spend attribution, model right-sizing, AI-fluency signals. Distributed as npx @promptster/cc-audit (a single-file bundle) and as standalone binaries.

Core principle: local-first, consent-tiered egress. The deterministic half (parse → attribute → report) runs fully local with no network and no key. Only explicit opt-in steps send anything, each gated proportional to what leaves the machine. Preserve this — never add a code path that phones home on a bare/non-interactive run.

Commands

pnpm build      # tsc → dist/  (run before node dist/cli.js or the bundlers)
pnpm typecheck  # tsc --noEmit
pnpm lint       # oxlint
pnpm test       # vitest run   (pnpm test:watch for watch)
pnpm dev        # tsc --watch

pnpm build:npm  # → bundles/npm/  (esbuild single file + clean package.json; the npm artifact)
pnpm bundle     # → bundles/      (bun --compile binaries + esbuild .mjs fallback; needs bun)

Package manager is pnpm, pinned by the packageManager field (pnpm-lock.yaml is the committed lockfile; pnpm install --frozen-lockfile in CI). scripts/only-pnpm.mjs runs on preinstall and REJECTS npm/yarn — this package is not published with that guard active elsewhere, but here it is the rule. Use corepack pnpm if a stale pnpm shadows the pinned one on PATH. Always pnpm build before running node dist/cli.js.

Architecture

src/cli.ts is the entry point and the only place that orchestrates I/O, prompts, and the consent flow. Everything it calls is a pure-ish module:

Ingest — adapters/claudeCode.ts reads ~/.claude/projects and adapters/codex.ts reads ~/.codex/sessions → model.ts Session/Span types. The model is tool-agnostic so a Cursor adapter can drop in later.

Codex is opt-in behind --codex, and that is a correctness choice, not caution. The rails do not observe the same things: Codex ships reasoning as encrypted_content (no thinkingChars), has no Read tool (no reads, so no redundant-read rate) and no plan mode. Those fields come back 0/[], which is indistinguishable from "measured and found absent" — averaged in silently, a Codex-heavy user's fluency signals fall because of what the FORMAT omits. Session.source exists so a signal can select its rail instead. Spend, tokens, models, tools and file ops are fully measured on both.

Read the full file on GitHub · 168 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.

7d ago First seen · 168 lines · 2,869 tokens per session scan A be9637eba7d6

Subscribe to this mod's changes

cc-audit CLAUDE.md is an instructions file published in the GitHub repository pa-arth/cc-audit (9 stars, last pushed 10d ago), licensed MIT. It adds 2,869 tokens to every session, about $0.0143 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 2026-08-31.

Analytics and metrics Claude Code ai-fluency anthropic cli cost-analysis developer-tools

next.js AGENTS.md

AGENTS.md instructions for vercel/next.js, covering next.js development guide, codebase structure, monorepo overview, core package: packages/next and other important packages.

vercel/next.js · 7,296 tokens

codex AGENTS.md

AGENTS.md instructions for openai/codex, covering rust/codex-rs, the codex-core crate, code review rules, crate api surface and model visible context.

openai/codex · 5,182 tokens

vscode buildNext.instructions.md

Working notes and architecture documentation for the new esbuild-based build system in build/next. Use when making changes to the new build pipeline (transpile/bundle commands, NLS plugin, source-map handling, resource copying, or self-hosting watch tasks).

microsoft/vscode · 6,785 tokens

cc-audit: Instructions file for Claude Code