AgentSkillsCN

optimize-plugin

当用户请求“验证插件”“优化插件”“检查插件质量”“审视插件结构”，或提及插件优化与验证相关任务时，可使用此技能。

SKILL.md
--- frontmatter
name: optimize-plugin
description: This skill should be used when the user asks to "validate a plugin", "optimize plugin", "check plugin quality", "review plugin structure", or mentions plugin optimization and validation tasks.
argument-hint: <plugin-path>
user-invocable: true
allowed-tools: ["Read", "Glob", "Bash(realpath *)", "Bash(bash:*)", "Task", "AskUserQuestion", "TaskCreate", "TaskUpdate"]

Plugin Optimization

Execute plugin validation and optimization workflow. Target: $ARGUMENTS

Background Knowledge

Template Compliance: Components MUST conform to templates in ${CLAUDE_PLUGIN_ROOT}/examples/. See references/template-validation.md for complete requirements (instruction-type/knowledge-type skills, agents).

Tool Patterns: See references/tool-patterns.md for invocation styles. Key: Skill/AskUserQuestion/TaskCreate require explicit "Use [tool] tool" phrasing.

Initialization

Use TaskCreate tool to track all four phases before starting work.

Phase 1: Discovery & Validation

Goal: Validate structure and detect issues. Orchestrator MUST NOT apply fixes.

Actions:

  1. •Resolve path with realpath and verify existence
  2. •Validate .claude-plugin/plugin.json exists
  3. •Validate components against ${CLAUDE_PLUGIN_ROOT}/examples/ templates
  4. •Assess architecture (ask about command migration if needed)
  5. •Run validation: python3 ${CLAUDE_PLUGIN_ROOT}/scripts/validate-plugin.py "$TARGET"
  6. •Compile issues by severity (Critical, Warning, Info)

Phase 2: Agent-Based Optimization

Goal: Launch agent to apply ALL fixes. Orchestrator does NOT make fixes directly.

Condition: Skip this phase if Phase 1 found no issues (proceed directly to Phase 4).

Actions:

  1. •Launch plugin-optimizer:plugin-optimizer agent with the following prompt content:
    • •Target plugin path (absolute path from Phase 1)
    • •Complete list of validation issues with severity levels
    • •Template compliance results from Phase 1
    • •Any user decisions from architecture assessment
  2. •Agent autonomously applies fixes (uses AskUserQuestion for template fix approvals)
  3. •Agent increments version in .claude-plugin/plugin.json after fixes (patch: fixes/optimizations, minor: new components, major: breaking changes)

Phase 3: Final Verification

Goal: Re-run validation to verify fixes.

Condition: Only execute if Phase 2 applied fixes.

Actions:

  1. •Re-run all validation scripts from Phase 1
  2. •Compare results with initial findings
  3. •If critical issues remain: resume agent with remaining issues
  4. •Document outcome (fixed vs remaining with rationale)

Phase 4: Final Deliverables

Goal: Deliver comprehensive optimization report and updated documentation to user.

Actions:

  1. •Generate complete validation report with all findings, fixes, and verification results
  2. •Produce component inventory summary showing plugin structure
  3. •Update README.md to accurately reflect current plugin state
  4. •See ./references/workflow-phases.md for detailed steps and ./references/report-template.md for format