This STDIO-only MCP server has 8 tools with minimal schema documentation and weak descriptions. Most tools lack parameter type definitions in visible schemas. Descriptions are present but generic (10-50 chars), falling below the 50-200 char baseline for LLM-optimized docs. No input validation guidance, error recovery hints, or output schema documentation visible. Tool names are reasonable (verb-noun style: look, go, talk_to, do, work, map, where, join) but parameter descriptions are sparse. The server is a narrative game interface, not a production API, which explains the informal structure, but even game tools need clear schemas for agent reasoning.

• Expand tool descriptions to 50-200 chars. For 'go': 'Move to an adjacent location by name. Returns the sensory description of the new place, including sights, sounds, NPCs, and available exits. Call map() first to see valid destinations.'
• Add enum constraints to string parameters. For 'go', return valid place names from map() and embed them as an enum in the schema. For 'do' and 'work', provide examples of valid actions (e.g., 'sit a deep hour', 'help with the nets') as an enum or pattern.
• Document output schemas explicitly. E.g., 'look returns {description: string, exits: string[], npcs: string[]}'. This enables agents to extract and chain data.
• Add error recovery hints to descriptions. E.g., 'go: If the place name is invalid, call map() to see available destinations, then retry.'
• Rename 'do' to 'perform_action' or 'attempt_activity' and add a 'valid_actions' discovery tool or include valid actions in the 'where' response.