Token Waste Audit — Security skill for Claude Code
Run the FR-26 token-waste audit — parse this project's Claude Code transcripts deterministically (requestId-deduped totals + leak-signature candidates, incl.
How to install Token Waste Audit
Installs to ~/.claude/commands/fusebase-dev-fusebase-flow-token-waste-audit.md
mkdir -p ~/.claude/commands && curl -fsSL https://raw.githubusercontent.com/fusebase-dev/fusebase-flow/HEAD/.claude/commands/token-waste-audit.md -o ~/.claude/commands/fusebase-dev-fusebase-flow-token-waste-audit.md Restart Claude Code, or start a new session, for it to be picked up.
What Token Waste Audit does
description: Run the FR-26 token-waste audit — parse this project's Claude Code transcripts deterministically (requestId-deduped totals + leak-signature candidates, incl. large-output), map findings to FR-26 rules, and report concrete fixes. Read-only; report goes to state/audit/ (gitignored). (FuseBase Flow)
/token-waste-audit
Invoke the **token-economy** skill's measurement path (`flow-skills/token-economy/SKILL.md` § Measure it).
- Run
python hooks/local/token-waste-audit.py(
Alternatives in Security
- Security Ownership Map — Map people-to-file ownership, compute bus factor, and identify risks 14.6k ★
- Skill Coverage Audit — Trace codepaths in diffs, map against tests, auto-generate missing coverage — use before shipping PRs 2.8k ★
- Review Artifact — Creates one polished, self-contained HTML artifact from review findings, audit notes, PR feedback, code review 652 ★
Full documentation available on GitHub
View Source RepositoryRelated Skills
09 Technical Roadmap
Parse Screaming Frog crawl, audit AI crawler access, SSR, IndexNow, schema; output prioritized technical roadm
Skills Suggest
Audit recent sessions for repeated tasks and AI-output-then-manual-verification patterns. Produces a ranked li
Claudemd Rules
Audit the HARD-rules manifest. Cross-references spec/hard-rules.json with rule-hits.jsonl bySection data over
Claudemd Sampling Audit
Retrospective batch scan of historical transcripts for 8 self-enforced HARD rules — text detectors (§10-V bann
Claude Revoke
Audit and revoke everything Claude Code has been given on Linux: trusted folders, "always allow" rules, saved
Audit Session
Find your own most-violated CLAUDE.md rules by auditing recent session transcripts
Related Agents
Deck Decomposer Agent
Turns a large body of unstructured material - a folder of documents, notes, transcripts, a knowledge base or a
Ndv Optimize
Performance optimization specialist. Use when code is slow, queries are expensive, bundle is large, or latency
Context Analyzer
Analyzes context usage patterns, identifies waste, and provides optimization recommendations. Use when the use