Personal reusable agent skills.

• staged-code-review: exhaustive iterative review of staged Git changes that fixes actionable issues while preserving the staged index.

• repo-fleet-standardization: multi-repository cleanup of docs, tooling config, shared sync templates, and GitHub metadata without touching application code.

bunx skills add hu553in/skills

Then invoke a skill by name when the workflow applies:

Use $staged-code-review to iteratively review and fix my staged changes, preserve my staged index exactly, and leave any new fixes unstaged.
Use $repo-fleet-standardization to standardize docs, tooling configs, sync templates, and GitHub metadata across related repositories without touching application code or staging/committing unrequested changes.

• <skill-name>/SKILL.md contains the reusable workflow.
• <skill-name>/agents/openai.yaml contains the OpenAI agent prompt metadata.
• skills.sh.json groups skills on the skills.sh repository page. It does not change CLI installation or any SKILL.md content.

When adding a skill:

1. Place it under <skill-name>/SKILL.md.
2. Add agent-specific metadata under <skill-name>/agents/ when needed.
3. Add the skill slug to skills.sh.json so the skills.sh page stays organized.