MCP server providing Code Mode execution for API discovery and querying across multiple bioinformatics APIs (REST, GraphQL, SPARQL). Supports sandboxed JavaScript/TypeScript execution with auto-staging of large responses into SQLite.
Server defines 6 tools with generally clear purposes but exhibits inconsistent schema documentation and parameter descriptions. Tools 1 - 2 (search/execute for REST APIs) and 4 - 5 (GraphQL/SPARQL variants) have detailed descriptions (150 - 250 chars) explaining discovery workflows and code execution patterns. However, input schemas are partially visible in source; Tool 3 (GraphQL execute) and Tool 5 (SPARQL execute) lack explicit parameter descriptions in the visible code snippet. Tool 6 (query_data) has minimal description (60 chars). Parameter types are declared (string, number) but many lack actionable constraints (e.g., no max_results bounds, no SQL injection guidance for query_data). Error handling is not visible in the provided source. Output schemas are not documented. Naming follows verb_noun convention (search, execute, query) but Tool 2 and Tool 3 share the name '{prefix}_execute', ambiguous for LLM selection when both are available. No tool annotations (readOnlyHint, destructiveHint) visible despite all tools being READ_ONLY.
Execute JavaScript code against the API in a sandboxed V8 isolate. Code runs with: api.get(path, params) for GET, api.post(path, body, params) for POST; searchSpec(query), listCategories(), getEndpoint(path) for API discovery; console.log() capture; auto-staging of large responses (>30KB) into SQLite with data_access_id for querying. Path params like /lookup/{id} are auto-interpolated from params. Remaining params become query string. Large datasets auto-stage for SQL queries.
Execute JavaScript code against a GraphQL endpoint in a sandboxed V8 isolate. Code runs with: gql.query(queryString, variables?) for GraphQL execution; schema.types(), schema.type(), schema.search() for introspection helpers; db.stage(), db.queryStaged() for staging; optional api.get/api.post for hybrid REST capability; console.log() capture. Optionally supports a second REST upstream for hybrid GraphQL+REST mode. Large responses (>30KB) auto-stage into SQLite.
Execute JavaScript code against a SPARQL endpoint in a sandboxed V8 isolate. Code runs with: sparql.query(queryString, opts?) for SELECT/CONSTRUCT/DESCRIBE; sparql.ask(queryString) for ASK queries returning boolean; sparql.raw(queryString, opts?) for unparsed JSON envelope; prefixes object with common ontology prefixes; prefixHeader pre-built PREFIX block; db.stage(), db.queryStaged() for staging; console.log() capture. Large responses (>30KB) auto-stage into SQLite.
Query staged data using SQL. Accepts a data_access_id from a previous auto-staged result and executes SQL queries against the SQLite tables created during staging. Returns query results with pagination support.
Search the API catalog (catalog mode) or OpenAPI spec (OpenAPI mode). Returns matching endpoints with full parameter documentation. Use this to discover API capabilities before calling {prefix}_execute. In catalog mode: accepts query/category/max_results parameters and performs keyword-based search over static ApiCatalog. In OpenAPI mode: accepts code parameter for agent-written JavaScript with full resolved OpenAPI spec and helper functions available.
Duplicate tool names: Two tools named '{prefix}_execute' (REST and GraphQL variants). LLMs cannot distinguish between them; both will appear in the tool list with identical names, forcing ambiguous selection.
Missing output schema documentation. Tools return complex objects (staged data with data_access_id, introspection results, query results) but no schema is provided. LLMs cannot plan downstream calls or extract required fields.
Insufficient parameter constraints. max_results lacks bounds (default 10, max 25 stated in description but not in schema). sql parameter in query_data has no length limit or injection guidance. code parameters accept arbitrary JavaScript with no validation hints.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 43 | 2026-07-28+ | v2 |
Search the GraphQL schema for types, fields, and operations. Returns matching schema elements with documentation. Use this to discover GraphQL capabilities before calling {prefix}_execute. Accepts code parameter for agent-written JavaScript with full introspection data and helper functions available.
Minimal description for query_data tool (60 chars). Does not explain when to use it, what data_access_id is, or how to construct SQL queries. Violates 10 - 1024 char guideline and lacks actionable context.
No error handling guidance visible. Tools accept code execution (JavaScript, GraphQL, SPARQL) but no documentation on error recovery, timeout behavior, or how LLMs should handle execution failures.