Loading ToolBench
Three tools with basic schemas and descriptions, but significant gaps in parameter documentation and output structure. ollama_chat has reasonable parameter coverage (prompt, system, model, temperature all described with constraints), but ollama_list_models and ollama_ping lack any input parameters and minimal context. No output schemas documented for any tool. Error handling returns plain text wrapped in isError flag, but lacks recovery guidance. Tool names follow verb_noun pattern (good), but descriptions are brief (avg 80 chars, below 194-char baseline). No tool annotations (readOnlyHint, etc.) despite all being read-only operations.
Send a prompt to Ollama and return the model response.
List models currently available from the configured Ollama server.
Check whether the configured Ollama server is reachable.
No output schemas documented. LLMs cannot predict response structure or plan downstream calls. ollama_chat returns {content: [{type, text}], isError?}, ollama_list_models returns formatted text, ollama_ping returns connection status, all undocumented.
ollama_list_models and ollama_ping have empty inputSchema ({}). While technically valid, descriptions lack context about when to call them or what they reveal. ollama_ping description does not explain it checks reachability; ollama_list_models does not mention it returns model names and sizes.
Error handling returns plain text error messages without recovery guidance. E.g., 'Ollama request failed (500): connection refused' tells LLM nothing about whether to retry, check configuration, or ask user. Missing pattern:recovery-guide.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | D | 59 | 2026-07-28+ | v2 |
No tool annotations despite all three tools being read-only. MCP spec (2026-07-28) supports readOnlyHint to signal safe operations. Absence forces LLM to infer safety from descriptions.
ollama_chat description does not document that it modifies no state (read-only) or that temperature parameter range 0 - 2 is Ollama-specific. Parameter descriptions lack format hints (e.g., 'model name as registered in Ollama, e.g. llama3.1:8b').