DevLens MCP

Instructions

Implementation Reference

• src/devlens/server.py:241-295 (handler)

  The core implementation of the 'get_server_docs' tool, which provides documentation for the MCP server based on the requested topic.

  def get_server_docs(topic: str = "overview") -> str:
      """Get documentation about the WebDocx MCP server.
  
      Provides guidance on server capabilities, tool usage, workflows, and best practices.
  
      Args:
          topic: Documentation topic - 'overview', 'tools', 'workflows', 'orchestration', or 'examples'
  
      Returns:
          Formatted documentation for the requested topic.
      """
      docs = {
          "overview": """
  # WebDocx MCP Server
  
  MCP server for intelligent web research. 12 tools in 3 layers.
  
  ## Tools
  Primitives: search_web, scrape_url, crawl_docs, summarize_page, extract_links
  Composed: deep_dive, compare_sources, find_related, monitor_changes
  Meta: suggest_workflow, classify_research_intent, get_server_docs
  
  ## Design
  - Composable: small tools combine powerfully
  - Smart: auto-orchestration via suggest_workflow
  - Efficient: Markdown output, token-optimized
  - Context-aware: workflows adapt to research state
  
  ## Usage
  search_web → scrape_url (simple)
  suggest_workflow (auto-recommends)
  deep_dive (multi-source aggregation)
  
  ## Topics
  tools, philosophy, workflows, orchestration, examples
  """,
          "tools": """
  # Tools
  
  ## Primitives (fast, focused)
  search_web(query, limit=5) - DuckDuckGo search, returns [{title,url,snippet}]
  scrape_url(url) - Extract clean Markdown with metadata
  summarize_page(url) - Headings only, triage before full scrape
  extract_links(url, filter_external=True) - Categorize internal/external links
  crawl_docs(root_url, max_pages=5) - Follow links, aggregate docs with TOC
  
  ## Composed (workflows)
  deep_dive(topic, depth=3) - Search + parallel scraping + aggregation
  compare_sources(topic, sources) - Analyze consensus/differences across 2-5 URLs
  find_related(url, limit=5) - Discover similar resources via content analysis
  monitor_changes(url, previous_hash) - Track content changes via hashing
  
  ## Meta (intelligence)
  suggest_workflow(query, known_urls=[]) - Auto-recommend optimal tool sequence
  classify_research_intent(query) - Detect research goal (7 patterns)

• src/devlens/server.py:240-240 (registration)

  Tool registration using the @mcp.tool() decorator.

  @mcp.tool()