Loading ToolBench
Tasqr MCP server — task state management for AI agents
Tasqr MCP has 5 tools with basic naming (verb_noun pattern) and descriptions present, but critical gaps undermine quality. Tool descriptions are adequate (50-100 chars) but parameter schemas are largely absent or incomplete. The create_tasks and update_tasks tools show input structure in the rubric but actual parameter descriptions are missing from visible code. No output schemas documented. Error handling guidance absent. Security-critical tool (handles encrypted task data) lacks permission gates, audit trail declarations, and error recovery patterns. STDIO transport caps protocol readiness at 50.
Claim the next available task from the queue with decrypted sensitive fields
Create one or more tasks with encrypted sensitive fields (title, description, metadata)
Retrieve one or more tasks by ID with decrypted sensitive fields
List tasks with optional filtering and pagination, with decrypted sensitive fields
Update one or more existing tasks with encrypted sensitive fields (title, description, metadata, output, note)
No input/output schemas visible for get_tasks, list_tasks, claim_next_task. Rubric requires documented schemas for LLM planning. Cannot verify parameter types or return structure.
create_tasks and update_tasks show array-of-objects structure but lack per-parameter descriptions (e.g., what does 'metadata' contain? what format?). Rubric requires every parameter to have a description.
No error handling guidance. Tools that modify encrypted task state (create_tasks, update_tasks, claim_next_task) must document recovery paths. E.g., 'If encryption fails, retry with fresh DEK' or 'If task locked, try claim_next_task instead.'
Inferred effective spec: <=2025-11-25.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | D | 51 | <=2025-11-25 | v2 |
No permission gates or scope declarations. Tools handle sensitive encrypted data but lack audit trail declarations or permission checks. Rubric requires 'read:tasks', 'write:tasks' scope declarations.
list_tasks lacks pagination parameters (limit, offset/cursor) and total count in output. Rubric baseline: 90% of production tools support pagination. Without it, large task lists blow context window.