agent-tasks-hooks

Agent Skills

A personal collection of skills for AI coding assistants — covering autonomous feature development, code review, DX/UX analysis, TDD, holistic debugging, and developer productivity.

Skills and agents follow the open Agent Skills format and work with Claude Code, Cursor, Codex, Gemini CLI, Copilot, Windsurf, OpenCode, and more.

New: The autonomous-workflow skill orchestrates end-to-end feature development through a phase-based pipeline with optional companion skills. See its dedicated section below.

Install

Tip — keep it tidy. Always pass --agent <your-tool> (e.g. --agent claude-code). Without it, npx skills symlinks every skill into ~24 different AI-tool directories at once (.codebuddy/, .continue/, .crush/, …). Scoping the install to the tool you actually use keeps your workspace clean and your git status short.

Recommended — Claude Code only:

npx skills add https://github.com/mthines/agent-skills \
  --skill confidence \
  --agent claude-code \
  --yes

To install all skills:

npx skills add https://github.com/mthines/agent-skills --all --agent claude-code

Universal (works with any Agent Skills-compatible tool — only use this if you switch between many tools):

npx skills add https://github.com/mthines/agent-skills --all

/plugin marketplace add mthines/agent-skills
/plugin install mthines-agent-skills@mthines

gemini extensions install https://github.com/mthines/agent-skills

git clone https://github.com/mthines/agent-skills.git ~/.cursor/skills/mthines-agent-skills

git clone https://github.com/mthines/agent-skills.git ~/.copilot/skills/mthines-agent-skills

git clone https://github.com/mthines/agent-skills.git ~/.agents/skills/mthines-agent-skills

git clone https://github.com/mthines/agent-skills.git ~/.agents/skills/mthines-agent-skills

What's Included

About skill loading. Each skill below is either agent-invokable or slash-command-only. The distinction matters for your token budget:

• Agent-invokable skills sit in your model's available-skills list every session — only the short description field, not the body. The model can invoke them via Skill() when it detects a matching task, without you typing /name.
• Slash-command-only skills (disable-model-invocation: true in their frontmatter) are not in the model's invokable list at all. They cost nothing in baseline context. They load only when you type /name or when another skill calls them via Skill() at runtime.

In all cases, the skill body (SKILL.md content + rules/) is loaded only on invocation, never automatically.

Orchestrators

Coordinate other skills to execute multi-step workflows. Agent-invokable.