Skill Usage Counter — Development skill for Claude Code
Measure which Claude Code/Codex skills you actually invoke.
How to install Skill Usage Counter
This entry records only its repository, not the path inside it, so there is no
exact command to give. Open momenbasel/skill-usage-counter and copy the folder into
~/.claude/skills/, or the file into ~/.claude/agents/.
What Skill Usage Counter does
Measure which Claude Code/Codex skills you actually invoke. Hook + backfill + report; privacy-safe.
Alternatives in Development
- Ccusage — CLI for analyzing Claude Code/Codex usage from local JSONL files 11.8k ★
- Codexbar — Show usage stats for OpenAI Codex and Claude Code, without having to login 8.8k ★
- Privacy Policy — Draft privacy policies with GDPR compliance considerations 7.8k ★
README
skill-usage-counter
Find out which Claude Code / Codex **skills you actually use** - and which of the hundreds installed have never once been invoked - so you can trim the library that loads into every session's context.
Why
Every installed skill's name and description load into the model's context at session start. A large library is a per-session token tax. Claude Code has no built-in usage counter, and the injected skill list means grepping transcripts naively gives false positives. This tool matches the exact tool-call signature, so counts are real.
What it does
- Live hook - a
PostToolUsehook records eachSkillinvocation as it happens. - Backfill - parses existing transcripts to seed historical counts.
- Report - top used, and installed-but-never-invoked (archive candidates).
- Export - emits a usage file other audit tools can consume.
Privacy
Stores only: skill name, count, first/last timestamp, and a salted SHA-1 hash of the project path (so "used in N projects" works without revealing which). Never stores prompts, tool args, file paths, or usernames. The data file is git-ignored.
Install
git clone https://github.com//skill-usage-counter.git
cp skill-usage-counter/scripts/counter.py ~/.claude/hooks/skill-usage-counter.py
# add the PostToolUse snippet from examples/settings-hook-snippet.json to ~/.claude/settings.json
python3 ~/.claude/hooks/skill-usage-counter.py backfill
python3 ~/.claude/hooks/skill-usage-counter.py report
Usage
python3 scripts/counter.py backfill # seed from ~/.claude/projects
python3 scripts/counter.py report --top 30 # ranked usage + never-used list
python3 scripts/counter.py export # machine-readable usage json
Environment overrides: `SKILL_USAGE_FILE`, `SKILL_USAGE_PROJECTS`, `SKILL_USAGE_SKILLS_DIR`, `SKILL_USAGE_SALT`.
Tests
pytest -q tests/
License
MIT
Related Skills
Tandem Skill
Persistent pair-programming between Claude Code and the OpenAI Codex CLI: one Codex thread per project, protoc
Claude Usage Report
Claude Code stop hook: per-prompt usage report with token costs, context window, and rate limits
Limit Lifeboat
Safe Claude Code and Codex account switching for Mac, with usage for every work and personal account.
Token Consumer
Measure and shrink the token cost of Claude Code skills, tier by tier: the always-on system-prompt tax, the pe
Benchmark Rerankers
Measure whether adding a reranker actually improves retrieval, by scoring reranked vs. un-reranked results on
Claimcheck Agents
List the coding agents claimcheck can measure, and what has actually been verified
Related Agents
Measure Response Time For Varying Prefix Match Lengths
import time from statistics import mean def time_compare(guess, iterations=1000): times = [] for _ in range(it
Performance Monitor
Monitor agent execution, track token usage, measure response quality, and optimize workflows
AI Evaluation Engineer
Measure AI quality honestly with datasets, metrics, and judges that survive scrutiny. Invoke for data-ai-team