Canhada-Labs

Audit Tokens — Security skill for Claude Code

Security community

Analyze audit-log.jsonl for ghost-token-waste patterns across the 6 PLAN-047 detectors.

How to install Audit Tokens

Installs to ~/.claude/commands/canhada-labs-ceo-orchestration-audit-tokens.md

Terminal

mkdir -p ~/.claude/commands && curl -fsSL https://raw.githubusercontent.com/Canhada-Labs/ceo-orchestration/HEAD/.claude/commands/audit-tokens.md -o ~/.claude/commands/canhada-labs-ceo-orchestration-audit-tokens.md

Restart Claude Code, or start a new session, for it to be picked up.

What Audit Tokens does

description: Analyze audit-log.jsonl for ghost-token-waste patterns across the 6 PLAN-047 detectors. Emits a markdown report (default) or JSONL stream; advisory-only (never blocks a session). argument-hint: "[window=30] [format=markdown|jsonl|json] [output=]"

--- K1 context fork: heavy analytic skill runs in a forked context (PLAN-135 W3 unit k1a) ---

context: fork

/audit-tokens — Token economy audit

Runs the 6 ghost-token detectors shipped in PLAN-047 Phase 1 (`retry_churn`,

Alternatives in Security

Agent Memory Vault — Markdown-first shared memory vault for Claude Code and Codex with SQLite, Zvec, Git, closeout, and audit 314 ★
Memory Self Review — Mine recent agent history (claude-mem + usage stats) for recurring failures and repeated patterns, audit MEMOR 297 ★
Audit Prune Completeness — Catch the github-mcp class of bug — CallTools handlers that exist but are invisible to Claude because ListTool 20 ★