panel: remove the Ollama card and its controller entry

The local 14B model is not usable from this machine (8.25 tok/s prompt
processing; pi sends a 9,510-token agent prompt against a 4,096 context, so it
was truncated to 2,050 and still needed ~4 min before the first token). Nothing
about the panel was worth keeping for it.

- portal/tools.py: drop the 'Ollama Local LLM' tile (no other references remain)
- services.toml: drop the ollama allowlist entry; 8 services remain
- service_controller.py + docs: keep the generic 'systemd-system' kind but drop
  the ollama-specific wording and mark it as currently unused

Verified on .13: /api/services no longer lists ollama, the deployed tools.py has
0 ollama matches, the other 8 cards still report correctly, portal is healthy.

portal/tools.py

@@ -85,9 +85,6 @@ def _cat() -> list[Tool]:
              "Chat admin console (admins only)", "https://pb.chat.lab.audasmedia.com.au", login=True, admin=True),
         Tool("ai-router", "AI Router", "ai",
              "OmniRoute LLM gateway dashboard", "https://omniroute.lab.audasmedia.com.au"),
-        Tool("ollama", "Ollama Local LLM", "ai",
-             "Local AI model server (Qwen 2.5 Coder on .27)", "http://192.168.20.27:11434",
-             service_id="ollama", ram_mb=9000, togglable=True, lan=True),
         Tool("n8n", "n8n", "ai",
              "Workflow automation", "https://n8n.lab.audasmedia.com.au", login=True),
         Tool("prefect", "Prefect", "ai",