Loading ToolBench
Exposes Raindrop.io bookmarks and collections to AI assistants via MCP
Three tools with complete schemas and descriptions, but significant gaps in parameter documentation and output schema clarity. Tool names follow verb_noun convention (list_, search_, get_). Descriptions are present (avg 120 chars) but lack actionable context for LLM selection. search_bookmarks has 6 parameters but only 3 have descriptions; list_collections and get_bookmark lack parameter descriptions entirely. No output schemas documented. Error handling is minimal, tool code returns plain text errors without recovery guidance. No tool annotations (readOnlyHint, idempotentHint) despite all being read-only operations.
Get a specific bookmark by ID
List all Raindrop.io collections (both root and nested)
Search Raindrop.io bookmarks with filters. Supports keyword search, tag filtering, and collection filtering.
Missing parameter descriptions in list_collections and get_bookmark. 'include_children' and 'raindrop_id' lack explanatory text; LLMs cannot infer intent from names alone.
No output schemas documented. Callers cannot predict response structure (fields, types, nesting). LLMs must infer from unstructured text output, risking parsing errors.
Error handling returns plain text without recovery guidance. 'Bookmark not found' gives no hint to call search_bookmarks or list_collections. Errors should guide next steps.
All three tools are read-only but lack tool annotations (readOnlyHint=true). 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 | 51 | <=2025-11-25 | v2 |
search_bookmarks returns up to 50 items per page but no guidance on result limits or pagination strategy. Large result sets bloat context; description should state typical result size and pagination behavior.