mcp-conductorlisted

# MCP Conductor — Multi-Server Orchestration ## Overview Teach your AI agent to orchestrate multiple MCP servers as a unified intelligence pipeline. Single MCP calls give shallow results. Chaining them gives 10x depth. ## When to Use - Task requires data from 2+ sources (search + scrape + analyze) - User asks to "research" something (implies multi-source) - Need to validate information across sources - Building a comparison or competitive analysis - Monitoring for changes over time - Any workflow where one tool's output feeds another tool's input ## Orchestration Patterns ### Pattern 1: Research Pipeline ``` TASK: "Research how company X handles authentication" Step 1: DISCOVER (Exa MCP) → web_search_exa: "company X authentication architecture blog" → Returns: 5-10 relevant URLs Step 2: EXTRACT (Bright Data MCP) → scrape_as_markdown: each URL from Step 1 → Returns: full page content as markdown Step 3: ANALYZE (Agent reasoning) → Read all scraped content → Extract: patterns, technologies, trade-offs → Cross-reference findings Step 4: ENRICH (GitHub MCP / gh CLI) → Search for open-source implementations mentioned → Read relevant source code for concrete examples Step 5: SYNTHESIZE (Agent output) → Combine all sources into structured analysis → Cite sources, compare approaches, recommend ``` ### Pattern 2: Competitive Intelligence ``` TASK: "Compare all tools that solve X" Step 1: SEARCH (Exa MCP — semantic search) → "tools libraries fra