{
  "schema": "vedokrok.public-item.v1",
  "release_id": "MHC-RPUB-20260920-75ad787a",
  "url": "/knowledge/retrieve-tool-schemas-just-in-time",
  "id": "MHC-D-RESEARCH-1168",
  "version": "0.1.0",
  "title": "Retrieve tool schemas just in time",
  "summary": "A tool can exist without occupying attention on every turn.",
  "kind": "protocol",
  "body": "Index tool descriptions and retrieve a small candidate set from the current intent. Present only those schemas as callable tools, while preserving a fallback when retrieval confidence is weak. Evaluate routing recall, final task success, latency and context cost together.",
  "limits": [
    "A router can hide the right tool. Optimize for end-task quality and safe fallback, not minimum schema count."
  ],
  "topics": [
    "union-ai-runtime-verification-2026"
  ],
  "intents": [],
  "source_ids": [
    "RS-ED013467FB90D730"
  ],
  "evidence": [
    {
      "claim": "Large tool catalogs can be routed semantically so only a request-relevant subset of tool schemas is placed in the model's active tool context.",
      "source_id": "RS-ED013467FB90D730",
      "role": "supports",
      "note": "Routing can hide the correct tool. Recall and fallback behavior must be evaluated alongside token cost, latency and tool-selection accuracy.",
      "locator": "9:19-24:27, retrieve tools, load schemas just in time, evaluate top-K and recover from misses"
    }
  ],
  "use_when": [
    "An agent has access to a large tool catalog and every request carries all tool descriptions into the model context."
  ],
  "avoid_when": [
    "A router can hide the right tool. Optimize for end-task quality and safe fallback, not minimum schema count."
  ],
  "example": "A finance question retrieves account and transaction tools instead of loading schemas for HR, calendar, CRM and deployment tools into the same request.",
  "check": "The system can report tool-retrieval recall and task success, not only the tokens saved by showing fewer schemas.",
  "steps": [
    "Index tools by the jobs and entities they operate on.",
    "Retrieve a bounded candidate set from the current request.",
    "Expose only those schemas to the model for the turn.",
    "Fall back to broader discovery when the router is uncertain.",
    "Measure missed-tool failures as well as cost and latency."
  ],
  "sources": [
    {
      "id": "RS-ED013467FB90D730",
      "title": "The 100-Tool Agent Is a Trap: Scaling with Semantic Routers and JIT Context",
      "url": "https://ai.engineer/talks/vh2VGuQ3zhY-100-tool-agent-is-trap-scaling-semantic"
    }
  ],
  "relations": [
    {
      "from": "MHC-D-RESEARCH-1168",
      "to": "MHC-D-RESEARCH-0707",
      "type": "useful_with",
      "url": "/knowledge/use-progressive-disclosure-for-agent-instructions"
    }
  ],
  "collections": [
    {
      "id": "RC-ACD6385FB70207D3",
      "title": "Operate AI agents as systems you can replay, verify and constrain",
      "url": "/collections/operate-ai-agents-as-systems-you-can-replay-verify-and-constrain"
    }
  ]
}