FogBugz MCP has 7 tools with explicit schemas and descriptions. Naming follows verb_noun convention (create_, update_, assign_, list_, search_, get_). Descriptions are present but generic (avg 80 chars, below 194-char baseline). Schemas are properly typed with required/optional fields. Critical gaps: no output schemas documented, no error handling guidance, no enum constraints for priority/status fields, no pagination support for list/search tools, and parameter descriptions lack format/range constraints. Security: credentials injected via env vars (good). Composition: tools are single-responsibility but lack chaining metadata (e.g., search_cases doesn't return project_id needed by create_case).

• Document output schemas for all tools. Example for fogbugz_create_case: 'Returns {case_id: number, case_url: string, created_at: ISO8601, project: string}'. This enables LLMs to chain calls and extract IDs.
• Add enum constraints for priority and status fields. Replace type: ['number','string'] with explicit enum: [1, 2, 3, 4, 5, 6, 7] or ['lowest', 'low', 'medium', 'high', 'highest', 'urgent']. Update descriptions to list valid values.
• Add pagination to list_my_cases and search_cases: add 'offset' (default 0) and 'limit' (default 20, max 100) parameters. Document in description: 'Returns max 20 results per call. Use offset to paginate. Response includes total_count.'
• Enhance error handling in tool descriptions. Add recovery guidance: 'If case not found, try search_cases() first. If permission denied, verify your FogBugz role. If API timeout, retry after 5 seconds.'
• Add format constraints to parameter descriptions. E.g., 'caseId: positive integer (1 - 999999)', 'assignee: email ([email protected]) or FogBugz username', 'attachmentPath: local file path, max 10MB, .png/.jpg/.pdf/.txt only'.

Inferred effective spec: <=2025-11-25.

• Relies on Logging (deprecated) - log to stderr or use OpenTelemetry