Task Objective
- This Skill is used to: create, manage, and optimize other skills, providing full lifecycle management from basic definitions to advanced composition
- Capabilities include: structured skill generation, introspection analysis, modular composition, iterative optimization, version control, safety alignment checks
- Trigger conditions: the user needs to create a new skill, analyze an existing skill, compose multiple skills, optimize skill design, or manage skill versions
Prerequisites
- Dependency description: the dependency packages and versions required by the scripts
pyyaml>=6.0
Operation Steps
- Standard process:
- Requirements analysis and context understanding
- The agent analyzes user goals, history, and constraints
- Combine with references/skill-standards.md to determine the skill type and specification
- Basic capability execution
- Generate skill definition: call
scripts/generate_skill.pyto generate a standardized YAML front matter and structure - Format validation: call
scripts/validate_skill.pyto check format compliance and safety rules
- Generate skill definition: call
- Advanced capability execution (as needed)
- Skill composition: call
scripts/compose_skills.pyto implement skill nesting, reuse, and concatenation - Version management: call
scripts/version_skill.pyto handle skill version iteration and difference comparison - Introspection analysis: the agent analyzes skill capability boundaries, optimization points, and composition possibilities
- Skill composition: call
- Optimization and iteration
- The agent adjusts the skill design based on feedback, referring to references/composition-patterns.md
- Execute safety alignment checks, referring to references/security-guidelines.md
- Requirements analysis and context understanding
- Optional branches:
- When creating a basic skill: execute steps 1-2
- When composing multiple skills: execute steps 1-3 (including composition)
- When optimizing an existing skill: execute steps 1-2-4 -When iterating versions: execute steps 1-2-3 (version management)-4
Resource Index
- Required scripts:
- scripts/generate_skill.py (Purpose and parameters: generate a standardized skill definition, accepting parameters such as skill_name, description, capabilities, dependencies)
- scripts/validate_skill.py (Purpose and parameters: validate skill format and safety rules, accepting parameters such as skill_definition, check_security=True)
- scripts/compose_skills.py (Purpose and parameters: compose multiple skills, accepting parameters such as skills_list, composition_mode)
- scripts/version_skill.py (Purpose and parameters: manage skill versions, accepting parameters such as action, skill_path, target_version)
- Domain references:
- references/skill-standards.md (When to read: refer to format specifications when generating skill definitions)
- references/composition-patterns.md (When to read: refer to composition patterns when composing skills)
- references/security-guidelines.md (When to read: refer to alignment rules when validating skill safety)
Notes
- Make full use of the agent's language understanding and reasoning capabilities for requirements analysis, creative design, and introspection evaluation
- Call scripts only for technical operations involving structured data processing, format validation, composition logic, etc.
- Content such as skill descriptions, process design, and optimization suggestions is generated directly by the agent and does not require script processing
- When composing skills, pay attention to module boundaries to avoid capability coupling
- Version management follows semantic versioning specifications and records the impact scope of each change
Usage Examples
- Basic Creation: The user says "create a data cleaning skill". The agent analyzes the requirements → calls generate_skill.py to generate a definition → the agent supplements the process description.
- Skill Composition: The user says "combine data processing and report generation". The agent analyzes the composition strategy → calls compose_skills.py → the agent adjusts the interface definition.
- Version Iteration: The user says "add a logging feature to this skill". The agent designs the change → calls version_skill.py to record the version → generates a new definition.