fix(ollama-cloud): capability-gate reasoning_effort + correct disable semantics

Three follow-up fixes to the salvaged reasoning_effort support, all verified
live against ollama.com /v1/chat/completions + /api/show on deepseek-v4-pro,
gemma3, and qwen3-coder:

1. Capability-gate on /api/show 'thinking'. The original ignored the
   supports_reasoning flag and emitted reasoning_effort for every model. Now
   gated: only models whose native /api/show capabilities list contains
   'thinking' (deepseek-v4 yes; gemma3 / qwen3-coder no) get reasoning_effort.
   Mirrors the LM Studio pattern — capability resolved once per (model,
   base_url) in run_agent._supports_reasoning_extra_body via a cached probe
   (hermes_cli.models.ollama_model_supports_thinking), threaded into the
   profile hook as supports_reasoning. No live HTTP in the per-request path.

2. Disable actually disables. Ollama Cloud defaults to thinking ON and IGNORES
   the extra_body.thinking:{type:disabled} shape (verified: still returned
   reasoning). The only working off switch is top-level reasoning_effort:'none'.
   The salvaged code returned ({}, {}) for enabled:false / effort:none, leaving
   thinking ON. Now emits {'reasoning_effort': 'none'}.

3. Omit unrecognized effort. The original forwarded any unknown string verbatim
   including 'minimal' (a real Hermes effort level). Ollama Cloud rejects
   unrecognized values with a hard HTTP 400 (accepted set: low/medium/high/
   max/none), so forwarding 'minimal' would break the request. Now omitted.

Core touches (run_agent.py, hermes_cli/models.py) add the capability probe;
the plugin profile only consumes the resolved flag. 24/24 profile tests green;
194 provider/transport tests unaffected.

hermes_cli/models.py

@@ -3306,6 +3306,54 @@ def lmstudio_model_reasoning_options(
     return []
 
 
+def ollama_model_supports_thinking(
+    model: str,
+    base_url: Optional[str],
+    api_key: Optional[str] = None,
+    timeout: float = 5.0,
+) -> Optional[bool]:
+    """Return True if an Ollama (Cloud or local) model advertises ``thinking``.
+
+    Probes the native ``/api/show`` endpoint and checks the ``capabilities``
+    list, which Ollama populates from the model's metadata (e.g.
+    ``deepseek-v4-pro`` → ``["completion", "tools", "thinking"]`` while
+    ``gemma3:27b`` → ``["completion", "vision"]``). This is the authoritative
+    capability source — the OpenAI-compat ``/v1/models`` endpoint omits it.
+
+    Returns:
+        True  — the model declares the ``thinking`` capability.
+        False — ``/api/show`` succeeded but the model has no ``thinking`` cap.
+        None  — the probe failed (unreachable / non-Ollama / error); the caller
+                decides the fallback (we treat None as "don't emit").
+    """
+    import httpx
+
+    server_url = (base_url or "").strip().rstrip("/")
+    if server_url.endswith("/v1"):
+        server_url = server_url[:-3]
+    if not server_url:
+        return None
+
+    bare_model = _strip_ollama_cloud_suffix((model or "").strip())
+    if not bare_model:
+        return None
+
+    token = str(api_key or "").strip()
+    headers = {"Authorization": f"Bearer {token}"} if token else {}
+
+    try:
+        with httpx.Client(timeout=timeout, headers=headers) as client:
+            resp = client.post(f"{server_url}/api/show", json={"name": bare_model})
+            if resp.status_code != 200:
+                return None
+            caps = resp.json().get("capabilities")
+            if isinstance(caps, list):
+                return "thinking" in caps
+    except Exception:
+        return None
+    return None
+
+
 def _fetch_github_models(api_key: Optional[str] = None, timeout: float = 5.0) -> Optional[list[str]]:
     catalog = fetch_github_model_catalog(api_key=api_key, timeout=timeout)
     if not catalog: