What problem does it solve? Modern polyglot codebases scatter code intelligence across separate language servers, making it hard to get unified go-to-definition, references, and hover data. This Skill guides the construction of a graph daemon (graphd) that aggregates multiple LSP clients into one consistent semantic graph with real-time incremental updates. ## Core Features & Use Cases - Multi-Language LSP Orchestration: Initialize and manage TypeScript, PHP, Go, Rust, and Python language servers concurrently with proper LSP 3.17 capability negotiation and lifecycle handling. - Unified Graph Schema: Transform LSP responses into nodes (files, symbols) and edges (contains, imports, calls, references) with strict consistency rules. - Semantic Index & Caching: Produce nav.index.jsonl navigation data, support LSIF import/export, and persist via SQLite/JSON cache layers. - Performance Contracts: Enforce targets like sub-100ms graph queries, sub-20ms cached symbol lookups, and 100k+ symbol scalability. - Use Case: You are building a code visualization platform for a monorepo mixing TypeScript and PHP. Use this Skill to design the LSP aggregator, graph construction pipeline, and WebSocket diff streaming that power live navigation. ## Quick Start Ask the agent to design a graphd server that aggregates TypeScript and PHP language servers into a unified symbol graph with real-time file-watcher updates.
