enrich_solution

Instructions

Implementation Reference

• src/longmem/server.py:486-526 (handler)

  MCP tool handler for enrich_solution. Receives entry_id and context, delegates to store.enrich_solution(), returns JSON with updated status or error.

  @mcp.tool()
  async def enrich_solution(
      entry_id: Annotated[
          str,
          Field(description="The id returned by save_solution, confirm_solution, or search_similar."),
      ],
      context: Annotated[
          str,
          Field(
              description=(
                  "New information that refines or extends the saved solution. "
                  "Write as a reusable insight: state the general pattern first, "
                  "then give specific details. E.g.: 'Port 4181 is used when 4180 "
                  "is already taken by another auth proxy in the same stack.'"
              )
          ),
      ],
  ) -> str:
      """
      Append new context to an already-saved solution.
  
      Call this when a conversation reveals additional details AFTER a solution
      was already saved — for example, a follow-up clarification that makes the
      solution more reusable across projects.
  
      This is NOT for failures (use add_edge_case for those). This is for
      enrichment: new facts, patterns, or context that improve the answer.
      """
      try:
          store, *_ = await _get_deps()
          try:
              await store.enrich_solution(entry_id, context)
          except ValueError as exc:
              return json.dumps({"updated": False, "error": str(exc)}, indent=2)
          return json.dumps({
              "updated": True,
              "id": entry_id,
              "message": "Solution enriched. Future retrievals will include the new context.",
          }, indent=2)
      except Exception as exc:
          return _db_error(exc)

• src/longmem/server.py:485-486 (registration)

  Registration of enrich_solution as an MCP tool via @mcp.tool() decorator on FastMCP instance.

  # ── tool: enrich_solution ────────────────────────────────────────────────────
  @mcp.tool()

• src/longmem/server.py:488-502 (schema)

  Input schema for enrich_solution: entry_id (str) and context (str) with Pydantic Field descriptions.

  entry_id: Annotated[
      str,
      Field(description="The id returned by save_solution, confirm_solution, or search_similar."),
  ],
  context: Annotated[
      str,
      Field(
          description=(
              "New information that refines or extends the saved solution. "
              "Write as a reusable insight: state the general pattern first, "
              "then give specific details. E.g.: 'Port 4181 is used when 4180 "
              "is already taken by another auth proxy in the same stack.'"
          )
      ),
  ],

• src/longmem/store.py:489-509 (helper)

  Storage-layer implementation: fetches the entry by ID, appends additional_context (separated by '---'), updates the LanceDB table, and re-syncs the FTS index.

  async def enrich_solution(self, entry_id: str, additional_context: str) -> None:
      """Append new context to an existing solution's text (not a failure — a refinement)."""
      sid = self._safe_id(entry_id)
      rows = await (
          self._table.query()
          .where(f"id = '{sid}'")
          .limit(1)
          .to_list()
      )
      if not rows:
          raise ValueError(f"Entry {entry_id!r} not found")
  
      current_solution: str = rows[0].get("solution") or ""
      updated_solution = current_solution + "\n\n---\n" + additional_context
  
      await self._table.update(
          updates={"solution": updated_solution, "updated_at": _now()},
          where=f"id = '{sid}'",
      )
      if self._fts:
          await self._fts_resync(entry_id)