Two tools with complete schemas and descriptions, but significant gaps in parameter documentation and output schema clarity. verify_decision and create_decision both have action/decision verbs and reasonable descriptions (100+ chars), meeting baseline naming and description standards. However, parameter descriptions lack specificity on constraints, formats, and valid ranges. Output schemas are not formally documented, responses are inferred from code. Error handling is present but generic. No tool annotations (readOnlyHint/destructiveHint) despite clear semantic differences (verify is read-only, create is destructive).

• Add formal output schema documentation to both tools. Document decision_id (string, UUID format), recommendation (enum: approve|reject|review), risk_score (number, 0-100), risk_level (enum: low|medium|high|critical), proof_hash (string, hex), verified (boolean), timestamp (ISO 8601 string).
• Replace free-form parameter descriptions with constraints. E.g., 'amount: positive number, 0.01 - 1,000,000 (max per policy)' and 'priority: one of high|critical' as enum in schema.
• Add tool annotations: verify_decision with readOnlyHint=true; create_decision with destructiveHint=true. This signals to agents which tools have side effects.
• Enhance error messages with recovery guidance. E.g., 'API key missing: set TRUSTOS_API_KEY environment variable' or 'Policy not found: available policies are [list]. Did you mean one of these?'
• Document metadata structure with examples. E.g., 'metadata: object with keys like amount (number), currency (string, ISO 4217), region (string), recipient_type (string). Example: {"amount": 50000, "currency": "USDC", "region": "US"}'.