Agent skill · writing content · erichowens
skill-creator
Guide for creating, improving, benchmarking, and packaging Claude Agent Skills (SKILL.md files). Invoke when users want to create a skill from scratch, improve or test an existing skill, benchmark skill performance with variance analysis, or optimize a skill description for triggering accuracy. Also invoke when users say "turn this into a skill", "make a skill for X", "help me write a SKILL.md", "my skill isn't firing correctly", or want to convert a workflow/conversation into a reusable skill. Invoke proactively when a conversation has produced a repeatable workflow worth capturing. If the user mentions SKILL.md, skill files, skill descriptions, or skill triggering, this skill applies.
Why this skill is useful
Adds executable scripts for benchmarking and optimizing skills that enhance the AI's ability to create and improve its own capabilities.
What it needs
Requires python installed locally. About 11k tokens when loaded. Last updated 2026-07-14. 180 stars on the source repository.
What this skill does
Skill Creator Create new skills and iteratively improve them. The lifecycle: 1. Decide what the skill should do and how 2. Draft the skill 3. Run test prompts through claude-with-the-skill 4. Evaluate results qualitatively and quantitatively with the user 5. Rewrite based on feedback 6. Repeat until satisfied, then expand the test set 7. Package and deliver Determine where the user is in this lifecycle and help them progress. Starting from scratch ("I want a skill for X") -- help narrow scope, draft, test, iterate. Already have a draft -- go straight to eval/iterate. Just want a better description -- jump to Description Optimization. Be flexible: if the user says "just vibe with me, no evals," do that. --- Available Resources Before starting, verify what is available in the environment. The full workflow requires these bundled resources: Resource Path Used for --- --- --- Grader instructions agents/grader.md Assertion evaluation Comparator instructions agents/comparator.md Blind A/B comparison Analyzer instructions agents/analyzer.md Benchmark pattern analysis Schema reference references/schemas.md evals.json, grading.json formats Assertion patterns references/eval-patterns.md Writing discriminating assertions Troubleshooting references/troubleshooting.md Common errors and fixes Eval review template assets/evalreview.html Description optimization UI Benchmark aggregator scripts/aggregatebenchmark python -m scripts.aggregatebenchmark Description optimizer scripts/runloop python -m scripts.runloop Eval viewer eval-viewer/generatereview.py Human review interface If resources are missing: The core loop (draft -> test -> review -> improve) still works without them -- grade inline and present results in-conversation instead of the browser viewer. Note what is unavailable to the user and adapt. --- Communicating with the user The skill creator serves people across a wide range of technical backgrounds. …
How to use it
Reference it in AdaL, Claude Code, Cursor or any coding agent — nothing to install:
@skills erichowens/skill-creator--d8a520