# Skill Optimizer

Skill Optimizer is a public Agent Skills lifecycle toolkit by hqhq1025.

Current release: v2.0.0.

It contains three installable `SKILL.md` packages:

- `skill-miner`: mines coding-agent histories, archived sessions, memories, summaries, and exported transcripts to discover repeated workflows that may deserve new skills.
- `skill-personalizer`: audits and tunes existing, downloaded, forked, or newly created skills for one user's local tools, paths, phrasing, workflows, and trigger behavior.
- `skill-generalizer`: converts local or private skills into safer, portable, public-ready skills for GitHub, marketplaces, teams, or cross-agent sharing.

Canonical repository:

- https://github.com/hqhq1025/skill-optimizer
- https://hqhq1025.github.io/skill-optimizer/

Important files:

- README.md: overview, install instructions, platform support, usage examples.
- README.zh-CN.md: Chinese overview and usage.
- docs/research-background.md: research and ecosystem background.
- skills/skill-miner/SKILL.md: skill for mining repeated workflows from agent history.
- skills/skill-personalizer/SKILL.md: skill for auditing and personalizing existing skills.
- skills/skill-generalizer/SKILL.md: skill for public release generalization.

Machine-readable resources:

- llms-full.txt: fuller AI-readable project context.
- repo-metadata.json: structured project facts.
- sitemap.xml: canonical public URLs.
- site/site-data.json: structured website facts.