Educational crypto portfolio analysis MCP server exposing read-only tools to Claude Code
Four tools with basic descriptions and empty/minimal input schemas. All tools are READ_ONLY and stateless, which is good for safety. However, descriptions are generic (10-50 chars), lack context on WHEN to use each tool, and parameter descriptions are absent. No output schemas documented. Naming is clear (verb_noun pattern), but the server lacks the depth expected of production tools. STDIO transport caps overall potential.
Analyze the bundled sample portfolio and return a compact summary.
Return static policy metadata for an asset symbol.
List simplified risk rules implemented in the codebase.
List asset symbols supported by the educational policy file.
Parameter descriptions missing. 'symbol' parameter in get_asset_policy has a description, but no other parameters across tools have descriptions. LLMs cannot infer parameter meaning from names alone.
Tool descriptions are too brief (10-50 chars). Lack context on WHEN to use each tool, what it returns, and how it relates to other tools. Baseline for A+ tools is 50-200 chars with explicit WHAT/WHEN/WHY.
Output schemas not documented. Tools return complex objects (dicts, lists of dicts) but no schema is provided. LLMs cannot plan downstream calls or extract fields reliably without knowing the response structure.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | D | 54 | 2026-07-28+ | v2 |
No error handling guidance. Tools raise ValueError (get_asset_policy) but provide no recovery hints. LLMs need actionable error messages: 'Unsupported asset symbol: BTC. Try list_supported_assets() to see valid symbols.'
No tool annotations (readOnlyHint, destructiveHint, idempotentHint). All tools are READ_ONLY and safe to retry, but this is not declared in the schema. Modern MCP servers should use tool annotations to signal safety properties.