Single tool 'list_groups' has a basic description but critically lacks input schema documentation and parameter descriptions. The tool accepts no parameters (empty input object), which is correctly reflected, but the description is minimal (58 chars, below the 194-char baseline). No output schema is documented, LLMs cannot infer the structure of the returned list[str]. No error handling guidance, no pagination support despite returning a potentially large list. The tool definition is visible in server.py via @mcp.tool() decorator, so it is not inferred.

• Add a detailed output schema: document that the tool returns a list of strings, each representing a group name (e.g., 'БПМ-21-1', 'ИВТ-22-2'). Include example values.
• Expand description to 150 - 250 characters: 'Returns all academic groups (учебные группы) at Moscow Polytechnic University. Use this to discover available groups before querying schedules or student rosters. Results are cached and updated hourly.'
• Add pagination support: accept optional 'limit' (default 50, max 200) and 'offset' (default 0) parameters. Return total_count so LLMs know if more results exist.
• Document error cases: 'If the cache is stale (>1 hour old), the tool will attempt to refresh from the API. If the API is unreachable, cached results are returned. If no cache exists, an error is raised with guidance to retry.'
• Rename to 'list_mospolytech_academic_groups' for clarity in multi-university contexts, or add a tool description prefix like 'Moscow Polytechnic University: List all academic groups.'