Open-Core Quantitative Finance Tools Provider - MCP server providing knowledge base search (RAG over research papers) and alpha vault telemetry submission for quantitative trading strategies
Both tools have basic descriptions and parameter schemas, but critical gaps undermine quality. search_knowledge_base lacks parameter descriptions entirely (violates pattern:tool-description). submit_alpha_telemetry has a vague description ('Submit...telemetry') that doesn't explain WHEN to use it or what 'successful' means. The 'metrics' parameter is documented as an object but lacks field-level schema detail (type, required fields, ranges). No output schemas are documented for either tool. Error handling is absent, no guidance on what happens if the knowledge base is empty, if telemetry submission fails, or how to recover. Tool names are acceptable (verb_noun pattern) but descriptions fall short of the 50-200 char LLM-optimized baseline (search_knowledge_base is 89 chars, acceptable; submit_alpha_telemetry is 68 chars but vague). No tool annotations (readOnlyHint, destructiveHint) despite submit_alpha_telemetry being a WRITE operation that should be marked destructive/irreversible.
Search the knowledge base of research papers using RAG (Retrieval-Augmented Generation)
Submit successful trading strategy telemetry to the Alpha Vault
search_knowledge_base: 'query' parameter has no description. LLMs cannot infer what kind of queries are valid (e.g., ticker symbols, paper titles, author names, research topics).
submit_alpha_telemetry: 'metrics' parameter is typed as 'object' with no field schema. LLMs cannot validate required fields (sharpe_ratio, total_return, max_drawdown, total_trades, win_rate) or their types/ranges. Description lists fields but schema does not enforce them.
No output schemas documented for either tool. LLMs cannot plan downstream calls or extract results. search_knowledge_base should document returned fields (e.g., paper_id, title, authors, relevance_score). submit_alpha_telemetry should document confirmation response (e.g., submission_id, status, timestamp).
Inferred effective spec: 2026-07-28+.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 38 | 2026-07-28+ | v2 |
submit_alpha_telemetry lacks error handling guidance. No description of what happens if submission fails (network error, validation error, duplicate strategy). LLMs have no recovery path.
submit_alpha_telemetry is a WRITE operation but has no tool annotation (destructiveHint). LLMs cannot distinguish it from read-only tools and may not understand the irreversible nature of telemetry submission.