Agent skill
skill-plan
[Tooling & Meta] Guide for creating effective skills, adding skill references, skill scripts or optimizing existing skills. This skill should be used when users want to create a new skill (or update an existing skill) that extends Claude's capabilities with specialized knowledge, workflows, frameworks, libraries or plugins usage, or API and tool integrations.
Install this agent skill to your Project
npx add-skill https://github.com/majiayu000/claude-skill-registry/tree/main/skills/other/other/skill-plan
SKILL.md
Skill Creator
Create and optimize Claude Code skills -- modular packages providing specialized workflows, tool integrations, and domain expertise.
Skill Structure
.claude/skills/skill-name/
SKILL.md # Required. <100 lines. Quick reference.
references/ # Docs loaded on-demand (also <100 lines each)
scripts/ # Executable code (Node.js/Python preferred)
assets/ # Output files (templates, images, fonts)
For detailed format spec, frontmatter fields, and file organization:
⚠️ MUST READ: .claude/skills/skill-plan/references/skill-structure-guide.md
Creation Workflow
- Understand -- Gather concrete usage examples; ask clarifying questions
- Plan -- Identify reusable resources (scripts, references, assets) from examples
- Initialize -- Run
scripts/init_skill.py <name> --path <dir>for scaffold - Implement -- Write SKILL.md (<100 lines), create references/scripts/assets
- Package -- Run
scripts/package_skill.py <path>to validate and zip
Constraints
SKILL.mdmust be <100 lines (frontmatter + body)- References/ files also <100 lines each; split further if needed
- Progressive disclosure: heavy content in references/, not SKILL.md
- Combine related topics into one skill (e.g., cloudflare-* -> devops)
- Imperative writing style ("Create X" not "You should create X")
- No "Task Planning Notes" boilerplate
- Scripts: prefer Node.js/Python, include tests, respect
.envloading order
Writing SKILL.md
- Purpose statement (1-2 lines)
- Decision tree or workflow steps (concise, <5 lines per step)
Read:directives to references/ for detailed guidance- No inline code examples >5 lines -- move to references/
Quality Gate
Before finalizing, validate against checklist:
⚠️ MUST READ: .claude/skills/skill-plan/references/skill-quality-checklist.md
Iteration
After using a skill on real tasks, watch for: Claude re-discovering known info, ignoring bundled scripts, false triggers, or missed triggers. Update SKILL.md, description keywords, or references accordingly.
IMPORTANT Task Planning Notes
- Always plan and break many small todo tasks
- Always add a final review todo task to review the works done at the end to find any fix or enhancement needed
Recommended Agent Skills
Expand your agent's capabilities with these related and highly-rated skills.
agent-ops-spec
Manage specification documents in .agent/specs/. Use when user provides requirements, acceptance criteria, or feature descriptions that need to be tracked and validated against implementation.
agent-ops-state
Maintain .agent state files. Use at session start, after meaningful steps, and before concluding: read/update constitution/memory/focus/issues/baseline consistently.
agent-ops-spec
Manage specification documents in .agent/specs/. Use when user provides requirements, acceptance criteria, or feature descriptions that need to be tracked and validated against implementation.
agent-ops-testing
Test strategy, execution, and coverage analysis. Use when designing tests, running test suites, or analyzing test results beyond baseline checks.
agent-ops-testing
Test strategy, execution, and coverage analysis. Use when designing tests, running test suites, or analyzing test results beyond baseline checks.
agent-ops-state
Maintain .agent state files. Use at session start, after meaningful steps, and before concluding: read/update constitution/memory/focus/issues/baseline consistently.
Didn't find tool you were looking for?