MCP server for interacting with Felo's LiveDoc API, enabling creation, management, and retrieval of documents, resources, tasks, and README content
MemClaw exposes 24 tools with basic schemas and descriptions, but falls short of production quality. All tools have descriptions (10-50 chars, mostly terse), and input schemas are present with type definitions. However, descriptions lack LLM-optimization guidance (e.g., 'when to use', 'what it returns'), parameter descriptions are minimal, output schemas are undocumented, and error handling provides no recovery guidance. No tool annotations (readOnlyHint, destructiveHint, idempotentHint) despite clear risk levels. Naming is verb-forward (create, list, update, delete, add, remove) but some tools conflate concerns (e.g., 'add-urls' + 'add-doc' could be unified as 'add-resource'). Parameter validation rules are absent from descriptions. No pagination guidance in list tools despite accepting page/size params.
Create text document in a LiveDoc
Add a comment to a task
Add URLs to a LiveDoc (max 10 URLs)
Append to README
Get text content of a resource
Create a LiveDoc
Create a task in a LiveDoc
Delete a LiveDoc
Descriptions are terse (10-50 chars) and lack LLM-optimization guidance. Missing 'when to use', 'what it returns', and dependency hints. Baseline for A+ tools is 50-200 chars with actionable context.
Output schemas are completely undocumented. LLMs cannot plan downstream tool calls or extract required fields (e.g., short_id, resource_id) without knowing response structure.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 46 | 2026-07-28+ | v2 |
Delete README
Delete a task
Download source file to disk
Get README (returns summary + content)
List LiveDocs
Delete a resource from a LiveDoc
Get a specific resource
List resources in a LiveDoc
List task records (comments + history)
List tasks in a LiveDoc
Update a LiveDoc
Create or replace README
Update resource title, snippet, or thumbnail
Update Markdown content of an ai_doc resource
Partially update a task
Upload file to a LiveDoc
No tool annotations (readOnlyHint, destructiveHint, idempotentHint) despite clear risk levels (READ_ONLY, WRITE, DESTRUCTIVE). Agents cannot distinguish safe from dangerous operations.
List tools (list, resources, tasks, task-records) accept page/size params but lack pagination guidance in descriptions. No mention of result limits, total counts, or next_cursor patterns.
Parameter descriptions are minimal or missing validation rules. E.g., 'status' in create-task accepts '0=TODO, 1=IN_PROGRESS, 2=DONE' but description lacks enum constraint or format guidance. LLMs may pass invalid values.