MCP server for detecting financial crime transaction patterns including structuring and layering, with risk scoring and SAR (Suspicious Activity Report) generation
Critical failures across all dimensions. The single tool 'execute_skill_action' has a vague, generic name that violates verb_noun conventions. Description is minimal (59 chars, below 10-char baseline). Input schema exists but lacks parameter descriptions entirely, transaction_sequence and account_id have no guidance on format, constraints, or usage. No output schema is documented. The implementation in mcp_server.py is a stub that prints JSON and exits; actual logic is in client.py (not integrated into MCP). No error handling, validation, or recovery guidance. Tool is not properly registered with the MCP server, it's merely listed in a JSON string. This server is non-functional as an MCP implementation.
Execute financial crime transaction pattern detection analysis
Tool name 'execute_skill_action' is generic and violates verb_noun convention. 'execute' and 'action' are vague, LLMs cannot infer intent. Should be 'detect_financial_crime_patterns' or 'analyze_transaction_patterns'.
Input parameters lack descriptions. 'transaction_sequence' and 'account_id' have no guidance on format, constraints, or expected values. LLMs cannot determine valid input without descriptions.
No output schema documented. The tool returns risk_score, flagged_patterns, sar_draft, and recommended_action, but LLMs have no way to know this structure exists or how to extract fields.
MCP server implementation is non-functional. mcp_server.py prints a static JSON string and exits; it does not implement the MCP protocol (no request/response handling, no tool invocation). Actual logic is in client.py, which is disconnected from the MCP server.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 21 | 2026-07-28+ | v2 |
No error handling or recovery guidance. If analysis fails or input is invalid, the tool provides no actionable error message or next steps for the LLM.