Query and analyze activity data with LLM agents. ActivityWatch tracks activities including foreground usage, time-on-task, and scheduled meetings; it cannot observe background apps, gauge productivity quality, or perform system configuration.
ActivityWatch MCP has 7 tools with consistent naming (all start with 'aw_get_' or 'aw_query_'), clear descriptions (100-150 chars each), and proper JSON Schema with typed parameters. However, parameter descriptions are minimal (10-30 chars), output schemas are not documented, and error handling guidance is absent. No tool annotations (readOnlyHint, destructiveHint) despite all being read-only. Schemas are present but lack depth, many parameters have no constraints (enums, min/max). Composition is sound (single responsibility per tool), but parameter naming could be more explicit (e.g., 'time_period' vs 'start'/'end' inconsistency across tools).
Get unified activity summary for a time period
Get calendar events for a specified time period
Get available buckets, capabilities, and suggested tools from ActivityWatch
Get context about current or upcoming meetings from calendar
Get a summary of activities for a specific time period
Get raw events from a specific ActivityWatch bucket
Execute a query against ActivityWatch data using the query language
Output schemas not documented. LLMs cannot plan downstream calls or extract fields without knowing response structure. All 7 tools lack documented return types.
Parameter descriptions are too brief (10-30 chars). 'Time period for activity query' lacks format guidance. Should specify: 'Time period as string (today, yesterday, this_week, custom_start/custom_end required for custom ranges)' to prevent LLM confusion.
No tool annotations despite all tools being read-only. Missing readOnlyHint in tool definitions. Agents cannot distinguish safe tools from destructive ones without explicit hints.
Inferred effective spec: <=2025-11-25.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | D | 55 | <=2025-11-25 | v2 |
Parameter constraints missing. 'limit' parameters lack min/max bounds; 'response_format' enums are present but other string params (bucket_id, query) accept free-form input without validation hints. No regex patterns or length constraints documented.
No error handling guidance. Tools lack recovery hints (e.g., 'If bucket not found, call aw_get_capabilities() to list available buckets'). Agents cannot self-correct on failures.