audit-config

A command that reviews your personal Claude Code instructions and rules for outdated, repetitive, or overly rigid guidance. It can check global settings, project settings, or both.

In plain words
What is it for?
Use it to review ~/.claude/CLAUDE.md, user-owned rule files, and project Claude Code configuration, then classify instructions that are outdated, redundant, or too prescriptive.
Why use it?
Long-lived instructions can waste context or tell current models to work around problems they no longer have. This helps identify guidance that may now reduce the quality of results.

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/cloudnative-co/claude-code-starter-kit/audit-config
Clone the repo
git clone --depth 1 https://github.com/cloudnative-co/claude-code-starter-kit
Per session 19 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 559 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. 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 $0.00019 $0.00559
Opus 5 $0.00010 $0.00280
Sonnet 5 $0.00004 $0.00112
Haiku 4.5 $0.00002 $0.00056

Measured 2d ago against content hash 3bb55042c967, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

audit-config 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 2d 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.

commands/audit-config.md · 56 lines

How it starts

The opening of the file, as written. The whole thing — 56 lines — stays where its author put it; the contents beside it link to each section on GitHub.

/audit-config — Personal Config Audit

Audit user-owned configuration for instructions that no longer help (or actively hurt) current-generation models. Kit-managed files are NOT in scope — the kit audits its own content upstream.

Scope

  • global (default): the user section of ~/.claude/CLAUDE.md (content outside the STARTER-KIT-MANAGED markers) and user-owned rule files (~/.claude/rules/user-*.md and any rules file not shipped by the kit)
  • project: the current project's CLAUDE.md / .claude/ config
  • all: both

Audit Lenses

Classify each instruction (line or section) into:

  1. babysitting — countermeasures for old-model failure modes that no longer occur (e.g. tool-call format reminders, "fix errors immediately", forced step-by-step rituals)
  2. over-prescription — rigid procedures that override the model's own judgment where it is now reliable (blanket TDD/coverage mandates, fixed command pipelines, arbitrary numeric gates)
  3. stale-premise — references to old models, retired features, or superseded mechanics (thinking keywords, old model names, obsolete context-management lore)
  4. context-tax — always-loaded lines duplicating what the harness or settings already enforce (language settings, parallel tool calls, risk reporting). Verify the duplication against the actual settings file or documented harness behavior before claiming it — cite what you checked.

Preservation Rules (do not over-delete)

  • Genuine personal preferences stay. Anything the model could not infer on its own — naming/secrecy rules, confirmation gates, response formats, workflow phrases — is the user's voice, not babysitting.
  • When unsure whether something is preference or babysitting, keep it and say why you were unsure.
  • Only mark context-tax removals when you actually verified the duplication (e.g. read settings.json, confirmed the harness ships the behavior). No speculation-based deletions.

Output (proposal only — never edit without approval)

Read the full file on GitHub · 56 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. 2d ago First seen · 56 lines · 19 tokens per session scan A 3bb55042c967

Subscribe to this mod's changes

audit-config is a command published in the GitHub repository cloudnative-co/claude-code-starter-kit (147 stars, last pushed 9d ago), licensed MIT. It adds 19 tokens to every session and 559 once invoked, about $0.0001 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.