knowledge-searching（知识搜索）

Knowledge Searching

Overview

Retrieves implementation knowledge to inform decision-making across the software development lifecycle.

Use this skill when you need:

• Implementation details for specific libraries/frameworks
• Code examples for patterns or features
• Documentation references libraries/frameworks usage

Announce at start: "I'm using the knowledge-research skill to gather implementation details."

Core Functions

Searching Specific Documentation:

1. Get sources → rag_get_available_sources() - Returns list with id, title, url
2. Find source ID → Match to documentation (e.g., "Supabase docs" → "src_abc123")
3. Search → rag_search_knowledge_base(query="vector functions", source_id="src_abc123")

General Research:

# Search knowledge base (2-5 keywords only!)
rag_search_knowledge_base(query="authentication JWT", match_count=5)

# Find code examples
rag_search_code_examples(query="React hooks", match_count=3)

Query Guidelines

✅ Good Queries (2-5 keywords)

• "authentication JWT"
• "vector functions"
• "React hooks"
• "Go context timeout"
• "SQL row level security"

❌ Bad Queries (too long/verbose)

• "How do I implement JWT authentication in Go?"
• "What are the best practices for vector similarity search?"
• "Show me examples of React hooks for state management"

Rule: Keep queries SHORT and keyword-focused for optimal search results.