feat(fabric): decommission Swapper from critical path, NCS = source of truth

- Node Worker: replace swapper_vision with ollama_vision (direct Ollama API)
- Node Worker: add NATS subjects for stt/tts/image (stubs ready)
- Node Worker: remove SWAPPER_URL dependency from config
- Router: vision calls go directly to Ollama /api/generate with images
- Router: local LLM calls go directly to Ollama /api/generate
- Router: add OLLAMA_URL and PREFER_NODE_WORKER=true feature flag
- Router: /v1/models now uses NCS global capabilities pool
- NCS: SWAPPER_URL="" -> skip Swapper probing (status=disabled)
- Swapper configs: remove all hardcoded model lists, keep only runtime
  URLs, timeouts, limits
- docker-compose.node1.yml: add OLLAMA_URL, PREFER_NODE_WORKER for router;
  SWAPPER_URL= for NCS; remove swapper-service from node-worker depends_on
- docker-compose.node2-sofiia.yml: same changes for NODA2

Swapper service still runs but is NOT in the critical inference path.
Source of truth for models is now NCS -> Ollama /api/tags.

Made-with: Cursor

services/node-worker/config.py

@@ -4,7 +4,6 @@ import os
 NODE_ID = os.getenv("NODE_ID", "noda2")
 NATS_URL = os.getenv("NATS_URL", "nats://dagi-nats:4222")
 OLLAMA_BASE_URL = os.getenv("OLLAMA_BASE_URL", "http://host.docker.internal:11434")
-SWAPPER_URL = os.getenv("SWAPPER_URL", "http://swapper-service:8890")
 DEFAULT_LLM = os.getenv("NODE_DEFAULT_LLM", "qwen3:14b")
 DEFAULT_VISION = os.getenv("NODE_DEFAULT_VISION", "llava:13b")
 MAX_CONCURRENCY = int(os.getenv("NODE_WORKER_MAX_CONCURRENCY", "2"))

services/node-worker/providers/ollama_vision.py

@@ -0,0 +1,49 @@
+"""Ollama vision provider — direct Ollama API with images, no Swapper dependency."""
+import logging
+from typing import Any, Dict, List, Optional
+
+import httpx
+
+from config import OLLAMA_BASE_URL, DEFAULT_VISION
+
+logger = logging.getLogger("provider.ollama_vision")
+
+
+async def infer(
+    images: Optional[List[str]] = None,
+    prompt: str = "",
+    model: str = "",
+    system: str = "",
+    max_tokens: int = 1024,
+    temperature: float = 0.2,
+    timeout_s: float = 60.0,
+) -> Dict[str, Any]:
+    model = model or DEFAULT_VISION
+
+    payload: Dict[str, Any] = {
+        "model": model,
+        "prompt": prompt or "Describe this image.",
+        "stream": False,
+        "options": {"num_predict": max_tokens, "temperature": temperature},
+    }
+    if images:
+        clean = []
+        for img in images:
+            if "," in img and img.startswith("data:"):
+                clean.append(img.split(",", 1)[1])
+            else:
+                clean.append(img)
+        payload["images"] = clean
+    if system:
+        payload["system"] = system
+
+    async with httpx.AsyncClient(timeout=timeout_s) as c:
+        resp = await c.post(f"{OLLAMA_BASE_URL}/api/generate", json=payload)
+        resp.raise_for_status()
+        data = resp.json()
+        return {
+            "text": data.get("response", ""),
+            "model": model,
+            "provider": "ollama_vision",
+            "eval_count": data.get("eval_count", 0),
+        }

services/node-worker/worker.py

@@ -9,7 +9,7 @@ from typing import Any, Dict
 import config
 from models import JobRequest, JobResponse, JobError
 from idempotency import IdempotencyStore
-from providers import ollama, swapper_vision
+from providers import ollama, ollama_vision
 import fabric_metrics as fm
 
 logger = logging.getLogger("node-worker")
@@ -27,9 +27,13 @@ async def start(nats_client):
     global _nats_client
     _nats_client = nats_client
 
+    nid = config.NODE_ID.lower()
     subjects = [
-        f"node.{config.NODE_ID.lower()}.llm.request",
-        f"node.{config.NODE_ID.lower()}.vision.request",
+        f"node.{nid}.llm.request",
+        f"node.{nid}.vision.request",
+        f"node.{nid}.stt.request",
+        f"node.{nid}.tts.request",
+        f"node.{nid}.image.request",
     ]
     for subj in subjects:
         await nats_client.subscribe(subj, cb=_handle_request)
@@ -160,7 +164,7 @@ async def _execute(job: JobRequest, remaining_ms: int) -> JobResponse:
             )
         elif job.required_type == "vision":
             result = await asyncio.wait_for(
-                swapper_vision.infer(
+                ollama_vision.infer(
                     images=payload.get("images"),
                     prompt=payload.get("prompt", ""),
                     model=model,
@@ -171,11 +175,20 @@ async def _execute(job: JobRequest, remaining_ms: int) -> JobResponse:
                 ),
                 timeout=timeout_s,
             )
+        elif job.required_type in ("stt", "tts", "image"):
+            return JobResponse(
+                job_id=job.job_id, trace_id=job.trace_id, node_id=config.NODE_ID,
+                status="error",
+                error=JobError(
+                    code="NOT_YET_IMPLEMENTED",
+                    message=f"{job.required_type} adapter coming soon; use direct runtime API for now",
+                ),
+            )
         else:
             return JobResponse(
                 job_id=job.job_id, trace_id=job.trace_id, node_id=config.NODE_ID,
                 status="error",
-                error=JobError(code="UNSUPPORTED_TYPE", message=f"{job.required_type} not implemented"),
+                error=JobError(code="UNSUPPORTED_TYPE", message=f"{job.required_type} not supported"),
             )
 
         logger.info(