List LLM Models (`list_llm_models`)

Return the list of available models for the current LLM provider (Ollama, Anthropic, or OpenAI‑compatible), so a client UI can offer a model picker before calling ai_chat with its optional model argument.

Usage

{
  "method": "tools/call",
  "params": {
    "name": "list_llm_models",
    "arguments": {}
  }
}

Parameters

None. The tool reads the same environment variables as get_llm_status.

Response

{
  "provider": "ollama",
  "models": [
    { "name": "llama3.2:latest" },
    { "name": "mistral:latest" }
  ]
}

• For "ollama", the list comes from GET {host}/api/tags and reflects the models actually installed on that server (empty if unreachable).
• For "anthropic" and "openai_compatible", a short hardcoded list of popular models is returned.
• For "none", models is empty.

Notes

• Takes no session_id — provider configuration is server‑wide, not per‑session.
• Read‑only; no chat or completion calls are made.