A Model Context Protocol server for Spec3 tools and services, providing access to race car documentation and reference materials for BMW E36 Spec3 racing builds
Three tools present with functional schemas but significant quality gaps. Tool descriptions are adequate (120-180 chars, within baseline) but lack actionable WHEN-to-use guidance. Schema definitions exist for all tools with proper JSON Schema structure, but parameter descriptions are minimal or absent. No input validation or error guidance in tool definitions. Parameter naming follows verb_noun convention (get_, list_) correctly. Composition is sound, tools are focused and chain-able. Missing: parameter constraints (enums), output schema documentation, error recovery guidance, and LLM-optimized descriptions that explain prerequisite calls and dependencies.
Get information about the user's 1994 BMW E36 325is Spec3 race car build. Returns current configuration, build status, modifications, and car-specific details. Call this tool when providing personalized advice, troubleshooting, or planning modifications.
Retrieve full text and visual content of Spec3 racing reference documents. Fetches complete PDF content from S3 including text and page images. Page images preserve diagrams, tables, and formatting that text extraction cannot capture.
List all available Spec3 racing reference documents. Available documents include: Spec3 Constructor's Guide, Bentley E36 Manual, 2025 NASA CCR rules, and 2025 Spec3 class rules.
Input schemas present but parameter descriptions missing or minimal. get_car_context and list_documents have empty input objects with no documentation of optional filters or behavior nuances.
Tool descriptions lack actionable WHEN-to-use context. 'Get information about the user's car' does not explain when to call this vs other tools, or what prerequisites exist. No guidance on retry behavior or error handling.
No output schema documentation. Tools return structured data but LLMs cannot see the expected response fields, forcing them to guess what data is available for downstream operations.
Inferred effective spec: <=2025-11-25.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-22 | F | 48 | <=2025-11-25 | v2 |
| 2026-03-09 | F | 0 | - | v1 |
get_document parameters (page_start, page_end, include_images) lack detailed format/range constraints in descriptions. page_start defaults to 1 but valid upper bound is not documented.
No error handling or recovery guidance in tool definitions. If document_id is invalid or S3 fetch fails, the tool definition provides no hint for LLM retry strategy.
Tool descriptions include implementation details (S3, PDF extraction) rather than user-facing abstractions. LLMs do not need to know about S3 or page image encoding, focus on capability and when to call.