MCP server for Federal Reserve statements and data analysis
FedSpeak has 4 tools with complete JSON schemas and basic descriptions, but falls short of production quality. All tools are READ_ONLY and well-named with action verbs (search_, get_, analyze_). Schemas include proper types and required fields. However, descriptions are generic (avg ~80 chars, below 194-char baseline), parameter descriptions lack actionable constraints (no enums, ranges, or format guidance), and output schemas are completely undocumented. Error handling is minimal, API errors return raw text without recovery guidance. No tool annotations (readOnlyHint, idempotentHint) despite all being read-only. The server proxies to a private API, so actual validation and error enrichment happen downstream, but the MCP layer provides no guardrails.
Analyze trends in Federal Reserve language over time
Get metadata about available FOMC statements
Get the most recent FOMC statement with full text
Search Federal Reserve (FOMC) statements semantically
Output schemas completely undocumented. LLMs cannot predict response structure, forcing them to guess at field names and types for downstream reasoning.
Parameter descriptions lack actionable constraints. 'Natural language query' and 'Topic to analyze' are vague; no guidance on valid topics, query length limits, or expected format. LLMs will hallucinate invalid values.
Error handling returns raw API text without recovery guidance. 'Error: The FOMC data service returned an error: <raw_text>' gives LLM no actionable next step. Should classify errors (retryable vs user-fixable) and suggest alternatives.
Inferred effective spec: <=2025-11-25.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 45 | <=2025-11-25 | v2 |
No tool annotations despite all tools being read-only. Missing readOnlyHint would help clients optimize caching and retry logic. Idempotent tools should declare idempotentHint.
Date parameters (start_date, end_date) lack format specification. Description says 'YYYY-MM-DD' but no JSON Schema format constraint. LLMs may pass invalid dates; no validation at MCP layer.