RSS MCP server has critical definition gaps. Tool descriptions are present but lack actionable detail for LLM selection. Parameter descriptions are minimal or absent (e.g., 'status' and 'source' in get_content lack guidance on valid values). Input schemas lack enum constraints, type specificity, and required field declarations. No output schemas documented. Error handling is absent, no guidance on recovery or invalid input handling. Naming is acceptable (verb-noun pattern) but parameter design forces LLMs to guess at constraints.

• Add enum constraint to 'status' parameter: {"type": "string", "enum": ["normal", "favorite"], "description": "Article status filter. 'normal' for unread/default articles, 'favorite' for starred/bookmarked articles."}
• Clarify 'source' parameter: change description to 'Feed source title (exact match). Pass empty string to retrieve articles from all subscribed feeds. Call get_sources() first to see available feed titles.' Add type validation to reject invalid feed names.
• Document output schemas for all three tools. Example for get_content: 'Returns array of article objects with fields: id (number), title (string), content (string), link (string), pubDate (ISO 8601), status (normal|favorite), source (string feed title).'
• Add 'required' array to input schemas. get_content: required=['status', 'limit']; set_tag: required=['articleId', 'status']; get_sources: required=[].
• Implement error handling with recovery hints. Example: 'Invalid articleId: {id}. Call get_content() to retrieve valid article IDs.' Return structured error objects with error_code, message, and suggested_action fields.