Agentled MCP Server

Instructions

Implementation Reference

• src/tools/knowledge.ts:114-134 (handler)

  Tool registration and handler definition for 'get_scoring_history'. It uses the client factory to call the underlying service.

      server.tool(
          'get_scoring_history',
          `Fetch scoring history for entities from the Knowledge Graph.
  Returns past scoring decisions (PROCEED_TO_IC, HOLD_FOR_REVIEW, REPOSITION, SCORED) with DMF scores and dates.
  Use this to see how entities were previously scored and calibrate future scoring runs.
  Returns both structured records and a compact text format for prompt injection.`,
          {
              entityName: z.string().optional().describe('Filter scoring history by entity name'),
              limit: z.number().min(1).max(500).optional().describe('Max records to return (default 100, max 500)'),
          },
          async ({ entityName, limit }, extra) => {
              const client = clientFactory(extra);
              const result = await client.getScoringHistory(entityName, limit);
              return {
                  content: [{
                      type: 'text' as const,
                      text: JSON.stringify(result, null, 2),
                  }],
              };
          }
      );

• src/client.ts:347-353 (handler)

  Client method implementation that performs the actual HTTP request to the API.

  async getScoringHistory(entityName?: string, limit?: number) {
      const query = new URLSearchParams();
      if (entityName) query.set('entityName', entityName);
      if (limit) query.set('limit', String(limit));
      const qs = query.toString();
      return this.request(`/knowledge/graph/scoring${qs ? `?${qs}` : ''}`);
  }