Loading ToolBench
Remember Me MCP server for managing contextual snippets, summaries, and rules
Server has 11 tools with visible schemas and descriptions, but quality is inconsistent. Naming follows verb_noun pattern (good), but descriptions are often generic or incomplete. No output schemas documented. Error handling returns dict with success/error fields but lacks recovery guidance. Tool composition is reasonable (separate get/list/set/remove), but parameter documentation is severely deficient. Average description length ~120 chars (below 194 baseline). Most parameters lack descriptions entirely.
Context for working with me This MUST always be loaded when working with me.
List rules for when working with me Args: context: Context the rule is stored within
Remove context rule for working with me
Get snippet for a context when working with me Args: context: Context key the snippet is stored within key: Key to reference the snippet with
List snippets for a context when working with me Args: context: Context key the snippet is stored within include_snippets: Include the full snippet content
Remove a snippet from the context for working with me SHOULD ONLY EVER BE USED TO REMOVE **SNIPPETS** SHOULD NOT BE USED TO REMOVE ANY OTHER TYPES Args: context: Context key to store snippet within key: Key to reference the snippet with
Parameter descriptions missing or minimal. 'context' param appears in 10 tools but descriptions are generic ('Context key...') without explaining what contexts are valid, format, or examples.
No output schemas documented. Tools return dict with 'success', 'data', 'error' fields, but the structure of 'data' is never specified. LLMs cannot plan downstream calls without knowing what fields to expect.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | D | 54 | 2026-07-28+ | v2 |
Set context snippet for working with me SHOULD ONLY EVER BE USED TO SAVE **SNIPPETS** SHOULD NOT BE USED TO SAVE ANY OTHER TYPES Store contextual snippets of code or text. Args: context: Context key to store snippet within mime_type: Mimetype of the snippet key: Key to reference the snippet with snippet: Content of the snippet **as string** - convert to string if required!
Get summary for a context when working with me Args: context: Context key the summary is stored within key: Key to reference the snippet with
List summaries for a context when working with me Args: context: Context key the summary is stored within include_summary: Include the full summary content
Remove a summary from the context for working with me SHOULD ONLY EVER BE USED TO REMOVE **SUMMARIES** SHOULD NOT BE USED TO REMOVE ANY OTHER TYPES Args: context: Context key to store summary within key: Key to reference the summary with
Set context summary for working with me SHOULD ONLY EVER BE USED TO SAVE **SUMMARIES** SHOULD NOT BE USED TO SAVE ANY OTHER TYPES Store contextual summaries of conversations or aspects thereof. Args: context: Context key to store summary within. Generally the format should be markdown. mime_type: Mimetype of the summary. Should be markdown or another text format as appropriate. key: Key to reference the summary with summary: Content of the summary
Error handling lacks recovery guidance. Errors return dict(success=False, error=str(e)) with raw exception messages. Try search_users() with a partial name.'
my_context_rule_remove has no description. Tool purpose is unclear, does it remove a single rule or all rules matching a pattern?
Destructive operations (snippet_remove, summary_remove, rule_remove) lack confirmation/dry-run pattern.