Private Home Assistant MCP service for ChatGPT Work, providing tools to control and monitor Home Assistant entities, sprinkler systems, solar edge systems, and other smart home devices
Two tools with minimal definitions. Both lack input schemas (empty {} objects), descriptions are present but generic (50-60 chars), and no parameter documentation exists. Tool names follow verb_noun pattern (get_, list_) but schemas are completely absent. No output schema documentation visible. Error handling not evident. This server fails basic definition quality standards.
Read CHEF iQ connection and setup state; never returns account credentials.
Read current food/ambient temperatures, units and freshness. Stale values are null.
No input schemas defined. Both tools declare Input: {} with no parameter definitions, types, or constraints.
No output schemas documented. Tool descriptions do not specify what fields are returned (e.g., get_chef_iq_status returns what? connection_status? setup_state? data types?). LLMs cannot plan downstream calls without knowing response structure.
Descriptions are generic and under 100 chars. 'Read CHEF iQ connection and setup state; never returns account credentials.' (60 chars) lacks context on when to call this vs list_chef_iq_probes, what the response structure is, or what 'setup state' means. Baseline for A+ tools: 194 chars average.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 34 | 2026-07-28+ | v2 |
No parameter descriptions. Both tools accept empty input {}, so no parameters to describe, but this also means tools cannot accept filters, options, or context. If list_chef_iq_probes should support pagination or filtering, parameters are missing entirely.
No pagination support visible. list_chef_iq_probes returns current probes but no limit, offset, page, or next_cursor parameters. If there are many probes, results could blow context window. Baseline: paginated tools accept limit and offset/cursor.