metadata.json

{
  "name": "schema-aware-routing",
  "description": "A pattern for using LLM-extracted metadata to route unstructured text into the correct database tables. Demonstrates multi-table writes to thoughts, people, interactions, and action_items based on structured fields extracted by an LLM from raw input.",
  "category": "recipes",
  "author": {
    "name": "Clay Dunker",
    "github": "claydunker-yalc"
  },
  "version": "1.0.0",
  "requires": {
    "open_brain": true,
    "services": ["OpenAI-compatible API (LLM + embeddings) — OpenAI direct or OpenRouter (canonical OB1 setup)"],
    "tools": ["Node.js 18+ or Deno"]
  },
  "tags": [
    "ingestion",
    "routing",
    "schema",
    "slack",
    "supabase",
    "llm",
    "metadata-extraction",
    "people-graph",
    "action-items"
  ],
  "difficulty": "intermediate",
  "estimated_time": "45 minutes",
  "created": "2026-03-19",
  "updated": "2026-03-19"
}