Quantified Claude — Development skill for Claude Code
Counts which Claude Code skills you actually reach for, plus the agents, models, and tokens behind them—read from the transcripts already on your disk.
How to install Quantified Claude
This entry records only its repository, not the path inside it, so there is no
exact command to give. Open kendrick/quantified-claude and copy the folder into
~/.claude/skills/, or the file into ~/.claude/agents/.
What Quantified Claude does
Counts which Claude Code skills you actually reach for, plus the agents, models, and tokens behind them—read from the transcripts already on your disk.
Alternatives in Development
- Tempest — Agentic Engineering that actually scales 161 ★
- Codex OpenAI Proxy — Proxy server to use ChatGPT Plus tokens with CLINE/Claude Code extensions via OpenAI API compatibility 124 ★
- Bm Stats — Show the basemind dashboard — resource footprint (disk + RAM) and activity (tool calls, per-tool histogram, es 94 ★
README
quantified-claude
Which Claude Code skills do you actually use? Not which ones you installed—which ones you _reach for_.
`quantified-claude` answers that, along with three more questions like it: which subagents you spawn, which models run where, and what all of it costs in tokens. Every number comes from the session transcripts already sitting on your disk, so there's nothing to instrument and nothing to wait for.
It reads those transcripts and writes two Markdown files. One is a Map of Content (MOC) of your skills, sorted by how often you've used them. The other is a usage report covering models, agents, and token spend. Drop them in an Obsidian vault or read them in any Markdown viewer.
Skills were the first thing it measured; agents, models, and tokens came next. The name is deliberately broad, so other facets can slot in later.
- What You Get
- How It Works
- Setup
- Everyday Use
- Keeping It Current
- A Note on Transcript Retention
- Privacy
- Development
- Contributing
- License
- Design
What You Get
Two files, regenerated side by side every time you render. Counts below are real, from one machine.
The Skills MOC
`~/.claude/skills/skill-usage-moc.md`, a map of every skill you own crossed with how much you use it.
## Used & Owned
| Skill | Uses | Tool | Slash | Tokens | Last used | Description |
|--------------------|-----:|-----:|------:|-------:|------------|----------------------------------|
| humanizer | 887 | 886 | 1 | 2.1B | 2026-09-15 | Rewrite AI-sounding text… |
| technical-writing | 326 | 326 | 0 | 198M | 2026-09-15 | Layered prose standard… |
| handoff | 230 | 8 | 222 | 224M | 2026-09-15 | Save a session handoff… |
## Owned but Never Use
Related Skills
Cogload
Keep your head cold. A thermometer for your Claude Code days: every hour scored 0 to 100 from the transcripts
Transcriptmeter
What your agent sessions cost, from the transcripts on your disk: tokens, cache hit ratio, peak context, list-
Zapara
Keep your head cold. How hard was today? zapara scores every hour you spent driving Claude Code, 0 to 100, fro
Rolepod Stats
Show rolepod evidence stats for this project — tier routes, verify/review verdicts, strong-dispatch overrides,
SessionSweep
Browse every agent session left on your disk, read the conversation, and delete it for real — transcript plus
Customer
Narrow a market down to one customer you can name, reach and describe by what they already do.
Related Agents
Media Analyst
Tracks what proven FPL managers and reputable analysts are actually doing - elite ownership as hard evidence,
Mer Goal Check
Independent verifier for the goal-check gate of /meridian-solve-task: checks a task's acceptance criteria agai
Agentille Plan Reviewer
Reviews a planner's draft plan BEFORE execution — checks the goal is right, the steps actually reach it, paral