This server has fundamental gaps in definition quality. Tool naming is generally acceptable (verb-first: get_, list_, put_, update_), but descriptions are sparse and lack LLM-optimization. Parameter schemas are present but lack type annotations and constraints. No output schemas are documented. Error handling is absent. The server defines 5 tools with varying quality, most lack depth needed for confident agent selection and execution.
Get the current date and day of the week
List all available projects and sections
List all tasks for the next N days (maximum of 14 days)
Create a new task.
Updates an existing task. Any field not specified will not be updated.
Missing output schema documentation. No tool declares what fields/structure it returns. LLMs cannot predict what data is available for downstream tool calls or extraction.
Parameter descriptions are minimal or missing type constraints. 'days' in list_tasks lacks min/max bounds (no mention of the 14-day cap in param description). 'date' in put_task and update_task lacks format specification, LLMs cannot know whether '2025-01-15' or 'next monday' or both are valid without explicit description.
Tool descriptions lack context on WHEN to use them and WHY. E.g., 'list_sections' description says 'List all available projects and sections' but does not explain when an agent should call this (discovery, before creating tasks) or what structure is returned.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-22 | D | 50 | 2026-07-28+ | v2 |
| 2026-03-09 | F | 38 | - | v1 |
No error handling guidance. Tools silently fail if API credentials are invalid, network fails, or invalid IDs are passed. Error messages should guide recovery (e.g., 'Invalid section_id. Call list_sections() to see available sections').
Naming ambiguity in 'put_task' vs 'update_task'. While verb-first is good, 'put' is HTTP jargon and less clear than 'create_task'. LLMs may conflate PUT (idempotent update) with CREATE (always new), leading to wrong tool selection.
No pagination or result limits documented. 'list_sections' and 'list_tasks' return entire datasets without documented caps. If a user has thousands of tasks or sections, responses could blow the context window. No mention of max result count or pagination strategy.
update_task performs a partial move operation without clear documentation. The logic ('if task.section_id != section_id') is buried in code; the parameter description does not explain this behavior or what happens if project_id/section_id are both specified.