skill-creator
Create new skills, modify and improve existing skills, and measure skill performance. Use when users want to create a skill from scratch, edit, or optimize an existing skill, run evals to test a skill, benchmark skill performance with variance analysis, or optimize a skill's description for better triggering accuracy.
Profile →npx skills add anthropics/skills --skill skill-creator --agent claude-code
Same command for any agent — swap --agent for codex, cursor, copilot.
Weekly change comes from our own snapshots, not the repository page — it measures attention, not adoption.
What it does
A skill for creating new skills, iterating on drafts, running evaluations, and benchmarking performance. It covers defining what a skill should do, drafting the skill, creating test prompts, running evaluations with Claude (via the ability to access the skill), and presenting qualitative and quantitative results. It also supports rewriting the skill based on user evaluation feedback, expanding the test set, and cycling through iterations until satisfactory. After finishing, it can trigger a separate script to optimize the triggering description.
How it works
- Decide the skill’s purpose and approximate how it should perform.
- Write a draft of the skill.
- Create test prompts and run claude-with-access-to-the-skill on them.
- Help evaluate results both qualitatively and quantitatively.
- Draft quantitative evals if missing; explain them to the user.
- Use the
eval-viewer/generate_review.pyscript to show results and metrics.
- Rewrite the skill based on user feedback and any clear flaws from benchmarks.
- Repeat until satisfied; expand test set and re-run at larger scale.
- If desired, run the separate skill description improver to optimize triggering.
When to use it
Use when creating a skill from scratch, editing or optimizing an existing skill, or running evals to test skill performance and triggering accuracy. Trigger guidance: use when the user wants to design or improve a skill, evaluate it with tests, or benchmark performance.
What it can touch
- Tools: claude-code
- Workflows: spawning with-skill and baseline runs, generating evals via
eval-viewer/generate_review.py, and managing an evaluation workspace structure as described in the workflow. - Outputs: iteration-based evaluation results, timing and token usage collected per run, and benchmark artifacts like
benchmark.jsonandbenchmark.md(produced by the aggregation script).
Caveats
- The review describes interactions with external scripts and evaluation workflows; exact outcomes depend on user-provided prompts and the results of the evaluations.
- The skill depicts a multi-step iterative process; failures or poor evaluation results require repeating steps and refining prompts, tests, and skill logic.
# Skill Creator A skill for creating new skills and iteratively improving them. At a high level, the process of creating a skill goes like this: - Decide what you want the skill to do and roughly how it should do it - Write a draft of the skill - Create a few test prompts and run claude-with-access-to-the-skill on them - Help the user evaluate the results both qualitatively and quantitatively - While the runs happen in the background, draft some quantitative evals if there aren't any (if there are some, you can either use as is or modify if you feel something needs to change about them). Then explain them to the user (or if they already existed, explain the ones that already exist) - Use the `eval-viewer/generate_review.py` script to show the user the results for them to look at, and also let them look at the quantitative metrics - Rewrite the skill based on feedback from the user's evaluation of the results (and also if there are any glaring flaws that become apparent from the quantitative benchmarks) - Repeat until you're satisfied - Expand the test set and try again at larger scale Your job when using this skill is to figure out where the user is in this process and then jump in
- Communicating with the user
- Creating a skill
- Capture Intent
- Interview and Research
- Write the SKILL.md
- Skill Writing Guide
- Writing Style
- Test Cases
- Running and evaluating test cases
- Step 1: Spawn all runs (with-skill AND baseline) in the same turn
- Step 2: While runs are in progress, draft assertions
- Step 3: As runs complete, capture timing data
- Step 4: Grade, aggregate, and launch the viewer
- What the user sees in the viewer
python -m scripts.aggregate_benchmark <workspace>/iteration-N --skill-name <name> nohup python <skill-creator-path>/eval-viewer/generate_review.py \ kill $VIEWER_PID 2>/dev/null python -m scripts.run_loop \ python -m scripts.package_skill <path/to/skill-folder>
What does the skill-creator skill do?
Create new skills, modify and improve existing skills, and measure skill performance. Use when users want to create a skill from scratch, edit, or optimize an existing skill, run evals to test a skill, benchmark skill performance with variance analysis, or optimize a skill's description for better triggering accuracy.
How do I install it?
Run `npx skills add anthropics/skills --skill skill-creator --agent claude-code` — it drops the skill into your project so the agent can pick it up. Swap the --agent value for codex, cursor or copilot if you use one of those.
Where does this skill come from?
From anthropics/skills, a repository with 166,205 stars. We read it straight from the repository tree rather than a submitted listing, so what you see here is what is actually published.
Is a popular skill a good skill?
Not necessarily. Stars measure attention, not adoption — a repository can trend for a week and be abandoned. That is why we show the weekly change from our own snapshots next to the total, instead of a single flattering number.