Unified MCP server for structured reasoning + execution tracking — one binary, four tool families
Critical gaps across naming, descriptions, and schemas. Tool names use wildcard patterns (think_*, ship_*, roadmap_*, signal_*) that are inferred rather than explicitly registered, cannot verify actual tool definitions. Only 4 of 9 tools have visible input schemas; 5 tools lack any schema documentation. Descriptions are present but generic (e.g., 'Structured reasoning tools family', under 50 chars, below the 194-char baseline). No output schemas documented. Error handling and recovery guidance absent. Composition issues: roadmap_export, roadmap_get, roadmap_set_status, tracker_setup have schemas but lack descriptions of what they return or when to use them.
Roadmap and project planning tools family
Export roadmap to markdown or JSON file
Retrieve roadmap chunks by ID with optional field projection
Set status of a roadmap chunk
Get current status of roadmap chunks
Execution tracking tools family for AI agents
Signal and notification tools family
Structured reasoning tools family for AI agents
Wildcard tool families (think_*, ship_*, roadmap_*, signal_*) lack explicit definitions. Tool names, actual parameter schemas, and return types are inferred, not visible in source. Cannot verify naming clarity, parameter descriptions, or output structure.
Generic, under-length descriptions for wildcard families. 'Structured reasoning tools family' (47 chars) and 'Execution tracking tools family' (31 chars) fall far below the 194-char baseline and do not explain WHAT each tool does, WHEN to use it, or what it returns. LLMs cannot distinguish between tools in the family.
No output schemas documented for any tool. LLMs cannot plan downstream calls or extract required fields. roadmap_export, roadmap_get, roadmap_status, tracker_setup have input schemas but no documented return types.
Inferred effective spec: 2025-06-18+.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 45 | 2025-06-18+ | v2 |
Set up issue tracker integration with consent elicitation
roadmap_status lacks description and schema. Tool name is clear but no documentation of what status values are valid, what the response contains, or when to call it instead of roadmap_get.
No error handling or recovery guidance. Tools do not document what errors are retryable, what user-fixable errors look like, or what the LLM should do if a call fails. No actionable error messages visible.