This is a routing layer that maps what you're trying to accomplish to MCP servers and APIs that can actually do it, with source tracing baked in. Think of it as a trust-aware directory: you describe an objective, and it suggests capability-matched candidates with provenance data. It's read-only, so you're querying an index rather than modifying anything. Reach for this when you're building systems that need to dynamically discover and vet which tools or integrations to invoke based on intent, especially in contexts where knowing the source and trust chain of a capability matters as much as the capability itself.
claude mcp add --transport stdio isometric-architect-ai-objective-index -- uvx ai-objective-index