ai-agent-skills

A personal collection of reusable AI Agent Skills for Codex, Claude Code, and local automation workflows. Each Skill is an independent subproject under skill/<skill-name>/. Generated outputs are git-ignored and never uploaded by default.

Skills

Skill	Path	Status	Purpose
origin-plot	skill/origin-plot/	v1.0.2 Documentation Map Patch on the v1.0.1 Core Execution Backend (agent onboarding refresh in v1.0.3)	Stable Origin / OriginPro plot execution backend for canonical data + YAML configs.

For setup, daily commands, the data-handling boundary with Codex, and the directory map, read skill/origin-plot/README.md. For agent-specific operating guides, read skill/origin-plot/AGENT_USAGE.md, skill/origin-plot/CODEX.md, and skill/origin-plot/CLAUDE.md.

Branch and tag conventions

• Active development happens on the skill branch.
• Feature releases tag as vX.Y-<skill>-<short-name> (e.g. v1.0-origin-plot-core-refactor); hardening or doc patches tag as vX.Y.Z-... (e.g. v1.0.1-origin-plot-core-archive-notes).
• Historical experiments (such as v0.9 smart-input) are preserved by immutable tags plus frozen archive branches; they are intentionally out of the v1.0+ mainline. See skill/origin-plot/archive/v0_9_smart_input_reference.md.

Contributing

See CONTRIBUTING.md for the smoke-test routine, pre-commit hygiene, and the rule that output/ must never be committed.