Bear MCP Server has critical definition quality gaps. Only 2 tools are exposed; both have minimal descriptions (16-11 chars respectively), which fall well below the 50-200 char baseline for LLM-optimized descriptions. Input schemas are present but severely underdocumented, parameters lack descriptions entirely. The list_notes tool accepts query and tag parameters, both marked nullable (type: ["string", null]), but provide no guidance on expected format, interaction, or constraints. The list_tags tool has empty input properties with no documentation. Tool names follow verb_noun pattern (list_*), which is good, but descriptions are too terse to guide LLM selection. No output schemas are documented in the source, and no error handling guidance is visible. The server appears to read from a Bear.app SQLite database (read-only), but this context is buried in code rather than surfaced in descriptions.
Search all notes
List tags
Tool descriptions are critically under-length (16 and 11 characters respectively). Baseline for A+ tools is 50-200 chars. Descriptions must explain WHAT the tool does, WHEN to use it, and any prerequisites so LLMs can select correctly.
Input parameters 'query' and 'tag' in list_notes lack descriptions. The schema shows type and null-ability but provides no guidance on format, valid values, interaction (AND vs OR), or examples. Parameter descriptions are mandatory for LLM understanding.
No documented output schemas. The source does not show what fields list_notes or list_tags return, their types, or which IDs downstream tools might need. LLMs cannot plan multi-step calls without knowing response structure.
Inferred effective spec: <=2025-11-25.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-22 | F | 35 | <=2025-11-25 | v2 |
| 2026-03-09 | F | 41 | 2025-03-26+ | v1 |
No error handling or recovery guidance. The source gives no indication of what errors are possible (database connection failure, invalid query syntax, no results) or what the LLM should do next.
Parameters 'query' and 'tag' are marked nullable (type: ["string", null]) but no description explains what null means or how it differs from omitting the parameter. Nullable types without documented semantics invite misuse.