Remove internal sd-server management, proxy to external sd-server

- Remove ProcessManager and process.py
- Add get_sd_server_url() config (env/config/default)
- Update routes to proxy to external sd-server URL
- Remove model switching (handled by external sd-server)
- Update CLI serve command
- Update tests for new architecture

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "tensors"
-version = "0.1.11"
+version = "0.1.12"
 description = "Read safetensor metadata and fetch CivitAI model information"
 readme = "README.md"
 requires-python = ">=3.12"

scripts/reinstall.py

@@ -55,9 +55,17 @@ def sync_to_junkpile() -> None:
     """Sync project to junkpile."""
     print("\n[3/4] Syncing to junkpile...")
     excludes = [
-        ".git", ".venv", "__pycache__", "*.pyc", ".mypy_cache",
+        ".git",
-        ".pytest_cache", ".ruff_cache", ".coverage", "*.egg-info",
+        ".venv",
-        "node_modules", ".tmp",
+        "__pycache__",
+        "*.pyc",
+        ".mypy_cache",
+        ".pytest_cache",
+        ".ruff_cache",
+        ".coverage",
+        "*.egg-info",
+        "node_modules",
+        ".tmp",
     ]
     cmd = ["rsync", "-avz", "--delete"]
     for exc in excludes:

tensors/cli.py

@@ -582,24 +582,22 @@ def reload(
 
 @app.command()
 def serve(
-    model: Annotated[str, typer.Option(help="Path to model file for sd-server.")],
     host: Annotated[str, typer.Option(help="Wrapper API listen address.")] = "127.0.0.1",
     port: Annotated[int, typer.Option(help="Wrapper API listen port.")] = 8080,
-    sd_port: Annotated[int, typer.Option(help="sd-server listen port.")] = 1234,
+    sd_server: Annotated[str | None, typer.Option(help="sd-server URL to proxy to.")] = None,
     log_level: Annotated[str, typer.Option(help="Log level.")] = "info",
 ) -> None:
-    """Start the sd-server wrapper API (transparent proxy with hot reload)."""
+    """Start the sd-server wrapper API (proxies to external sd-server)."""
     try:
         import uvicorn  # noqa: PLC0415
 
-        from tensors.server import ServerConfig, create_app  # noqa: PLC0415
+        from tensors.server import create_app  # noqa: PLC0415
     except ImportError:
         console.print("[red]Missing server dependencies. Install with:[/red]")
         console.print("  pip install tensors[server]")
         raise typer.Exit(1) from None
 
-    config = ServerConfig(model=model, port=sd_port)
+    uvicorn.run(create_app(sd_server_url=sd_server), host=host, port=port, log_level=log_level)
-    uvicorn.run(create_app(config), host=host, port=port, log_level=log_level)
 
 
 # =============================================================================

tensors/config.py

@@ -243,3 +243,34 @@ def set_default_remote(name: str | None) -> None:
     else:
         config["default_remote"] = name
     save_config(config)
+
+
+# ============================================================================
+# SD Server Configuration
+# ============================================================================
+
+SD_SERVER_DEFAULT_URL = "http://localhost:1234"
+
+
+def get_sd_server_url() -> str:
+    """Get the sd-server URL.
+
+    Resolution order:
+    1. SD_SERVER_URL environment variable
+    2. config.toml [server].sd_server_url
+    3. Default: http://localhost:1234
+    """
+    # Check environment variable first
+    env_url = os.environ.get("SD_SERVER_URL")
+    if env_url:
+        return env_url
+
+    # Check config file
+    config = load_config()
+    server_config = config.get("server", {})
+    if isinstance(server_config, dict):
+        url = server_config.get("sd_server_url")
+        if url:
+            return str(url)
+
+    return SD_SERVER_DEFAULT_URL

tensors/server/__init__.py

@@ -1,4 +1,4 @@
-"""sd-server wrapper — FastAPI app for managing and proxying to sd-server."""
+"""sd-server wrapper — FastAPI app for proxying to an external sd-server."""
 
 from __future__ import annotations
 
@@ -12,42 +12,39 @@ from fastapi import FastAPI
 from fastapi.responses import FileResponse
 from fastapi.staticfiles import StaticFiles
 
+from tensors.config import get_sd_server_url
 from tensors.server.civitai_routes import create_civitai_router
 from tensors.server.db_routes import create_db_router
 from tensors.server.download_routes import create_download_router
 from tensors.server.gallery_routes import create_gallery_router
 from tensors.server.generate_routes import create_generate_router
-from tensors.server.models import ServerConfig
 from tensors.server.models_routes import create_models_router
-from tensors.server.process import ProcessManager
 from tensors.server.routes import create_router
 
 if TYPE_CHECKING:
     from collections.abc import AsyncIterator
 
-__all__ = ["ProcessManager", "ServerConfig", "app", "create_app"]
+__all__ = ["app", "create_app"]
 
 logger = logging.getLogger(__name__)
 
 
-def create_app(config: ServerConfig | None = None) -> FastAPI:
+def create_app(sd_server_url: str | None = None) -> FastAPI:
-    """Build the FastAPI application with process manager and proxy client."""
+    """Build the FastAPI application that proxies to an external sd-server.
-    pm = ProcessManager()
+
+    Args:
+        sd_server_url: URL of the sd-server to proxy to. If None, uses
+                       get_sd_server_url() to resolve from env/config.
+    """
+    backend_url = sd_server_url or get_sd_server_url()
 
     @asynccontextmanager
     async def lifespan(_app: FastAPI) -> AsyncIterator[None]:
+        _app.state.sd_server_url = backend_url
+        logger.info(f"Proxying to sd-server at: {backend_url}")
         async with httpx.AsyncClient(timeout=300) as client:
             _app.state.client = client
-            if config is not None:
-                pm.start(config)
-                logger.info("waiting for sd-server to become ready...")
-                ready = await pm.wait_ready()
-                if ready:
-                    logger.info("sd-server is ready")
-                else:
-                    logger.warning("sd-server did not become ready in time")
             yield
-        pm.stop()
 
     app = FastAPI(title="sd-server wrapper", lifespan=lifespan)
 
@@ -68,11 +65,10 @@ def create_app(config: ServerConfig | None = None) -> FastAPI:
     app.include_router(create_civitai_router())  # Must be before catch-all proxy
     app.include_router(create_db_router())  # Must be before catch-all proxy
     app.include_router(create_gallery_router())  # Must be before catch-all proxy
-    app.include_router(create_models_router(pm))  # Must be before catch-all proxy
+    app.include_router(create_models_router())  # Must be before catch-all proxy
     app.include_router(create_download_router())  # Must be before catch-all proxy
-    app.include_router(create_generate_router(pm))  # Must be before catch-all proxy
+    app.include_router(create_generate_router())  # Must be before catch-all proxy
-    app.include_router(create_router(pm))
+    app.include_router(create_router())
-    app.state.pm = pm
     return app
 
 

tensors/server/generate_routes.py

@@ -5,18 +5,15 @@ from __future__ import annotations
 import base64
 import logging
 import time
-from typing import TYPE_CHECKING, Any
+from typing import Any
 
 import httpx
-from fastapi import APIRouter, HTTPException
+from fastapi import APIRouter, HTTPException, Request
 from pydantic import BaseModel as PydanticBaseModel
 from pydantic import Field
 
 from tensors.server.gallery import Gallery
 
-if TYPE_CHECKING:
-    from tensors.server.process import ProcessManager
-
 logger = logging.getLogger(__name__)
 
 
@@ -113,38 +110,30 @@ def _process_image(
     return image_info
 
 
-def _check_server_running(pm: ProcessManager) -> None:
-    """Check if sd-server is running, raise HTTPException if not."""
-    if pm.proc is None or pm.proc.poll() is not None:
-        raise HTTPException(status_code=503, detail="sd-server is not running")
-    if pm.config is None:
-        raise HTTPException(status_code=503, detail="sd-server not configured")
-
-
 # =============================================================================
 # Router Factory
 # =============================================================================
 
 
-def create_generate_router(pm: ProcessManager) -> APIRouter:
+def create_generate_router() -> APIRouter:
     """Build a router with /api/generate endpoint."""
     router = APIRouter(prefix="/api", tags=["generate"])
     gallery = Gallery()
 
     @router.post("/generate")
-    async def generate(req: GenerateRequest) -> dict[str, Any]:
+    async def generate(request: Request, req: GenerateRequest) -> dict[str, Any]:
         """Generate images with gallery integration."""
-        _check_server_running(pm)
+        sd_server_url = request.app.state.sd_server_url
-        assert pm.config is not None  # Verified by _check_server_running
-
         body = _build_sd_request(req)
-        url = f"http://127.0.0.1:{pm.config.port}/sdapi/v1/txt2img"
+        url = f"{sd_server_url}/sdapi/v1/txt2img"
 
         try:
             async with httpx.AsyncClient(timeout=300) as client:
                 response = await client.post(url, json=body)
                 response.raise_for_status()
                 result = response.json()
+        except httpx.ConnectError as e:
+            raise HTTPException(status_code=503, detail=f"Cannot connect to sd-server: {e}") from e
         except httpx.HTTPError as e:
             logger.exception("Generation failed")
             raise HTTPException(status_code=502, detail=f"sd-server error: {e}") from e
@@ -153,8 +142,11 @@ def create_generate_router(pm: ProcessManager) -> APIRouter:
         info = _parse_info(result.get("info", {}))
         all_seeds = info.get("all_seeds", [req.seed] * len(images_data))
 
+        # Get model info from sd-server response if available
+        model_name = info.get("sd_model_name") or info.get("model")
+
         output_images = [
-            _process_image(img_b64, i, all_seeds[i] if i < len(all_seeds) else req.seed + i, req, gallery, pm.config.model)
+            _process_image(img_b64, i, all_seeds[i] if i < len(all_seeds) else req.seed + i, req, gallery, model_name)
             for i, img_b64 in enumerate(images_data)
         ]
 
@@ -167,32 +159,34 @@ def create_generate_router(pm: ProcessManager) -> APIRouter:
         }
 
     @router.get("/samplers")
-    async def list_samplers() -> dict[str, Any]:
+    async def list_samplers(request: Request) -> dict[str, Any]:
         """List available samplers from sd-server."""
-        _check_server_running(pm)
+        sd_server_url = request.app.state.sd_server_url
-        assert pm.config is not None
+        url = f"{sd_server_url}/sdapi/v1/samplers"
-        url = f"http://127.0.0.1:{pm.config.port}/sdapi/v1/samplers"
 
         try:
             async with httpx.AsyncClient(timeout=30) as client:
                 response = await client.get(url)
                 response.raise_for_status()
                 return {"samplers": response.json()}
+        except httpx.ConnectError as e:
+            raise HTTPException(status_code=503, detail=f"Cannot connect to sd-server: {e}") from e
         except httpx.HTTPError as e:
             raise HTTPException(status_code=502, detail=f"sd-server error: {e}") from e
 
     @router.get("/schedulers")
-    async def list_schedulers() -> dict[str, Any]:
+    async def list_schedulers(request: Request) -> dict[str, Any]:
         """List available schedulers from sd-server."""
-        _check_server_running(pm)
+        sd_server_url = request.app.state.sd_server_url
-        assert pm.config is not None
+        url = f"{sd_server_url}/sdapi/v1/schedulers"
-        url = f"http://127.0.0.1:{pm.config.port}/sdapi/v1/schedulers"
 
         try:
             async with httpx.AsyncClient(timeout=30) as client:
                 response = await client.get(url)
                 response.raise_for_status()
                 return {"schedulers": response.json()}
+        except httpx.ConnectError as e:
+            raise HTTPException(status_code=503, detail=f"Cannot connect to sd-server: {e}") from e
         except httpx.HTTPError as e:
             raise HTTPException(status_code=502, detail=f"sd-server error: {e}") from e
 

tensors/server/models.py

@@ -2,16 +2,5 @@
 
 from __future__ import annotations
 
-from pydantic import BaseModel
+# Note: ServerConfig and ReloadRequest were removed since we no longer manage
-
+# sd-server processes internally. The wrapper now proxies to an external sd-server.
-DEFAULT_PORT = 1234
-
-
-class ReloadRequest(BaseModel):
-    model: str
-
-
-class ServerConfig(BaseModel):
-    model: str
-    port: int = DEFAULT_PORT
-    args: list[str] = []

tensors/server/models_routes.py

@@ -3,30 +3,18 @@
 from __future__ import annotations
 
 import logging
-from pathlib import Path
 from typing import TYPE_CHECKING, Any
 
-from fastapi import APIRouter, HTTPException
+from fastapi import APIRouter, Request
-from fastapi.responses import JSONResponse
-from pydantic import BaseModel as PydanticBaseModel
 
 from tensors.config import MODELS_DIR
 
 if TYPE_CHECKING:
-    from tensors.server.process import ProcessManager
+    from pathlib import Path
 
 logger = logging.getLogger(__name__)
 
-
+_HTTP_OK = 200
-# =============================================================================
-# Request/Response Models
-# =============================================================================
-
-
-class SwitchModelRequest(PydanticBaseModel):
-    """Request body for switching models."""
-
-    model: str  # Path to model file
 
 
 # =============================================================================
@@ -76,7 +64,7 @@ def scan_checkpoints(directory: Path | None = None) -> list[dict[str, Any]]:
 # =============================================================================
 
 
-def create_models_router(pm: ProcessManager) -> APIRouter:
+def create_models_router() -> APIRouter:
     """Build a router with /api/models/* endpoints."""
     router = APIRouter(prefix="/api/models", tags=["models"])
 
@@ -90,62 +78,34 @@ def create_models_router(pm: ProcessManager) -> APIRouter:
         }
 
     @router.get("/active")
-    def get_active_model() -> dict[str, Any]:
+    async def get_active_model(request: Request) -> dict[str, Any]:
-        """Get information about the currently loaded model."""
+        """Get information about the currently loaded model from sd-server."""
-        status = pm.status()
+        import httpx  # noqa: PLC0415
-        config = pm.config
+
+        sd_server_url = request.app.state.sd_server_url
+
+        # Try to get current model from sd-server's options endpoint
+        try:
+            async with httpx.AsyncClient(timeout=10) as client:
+                response = await client.get(f"{sd_server_url}/sdapi/v1/options")
+                if response.status_code == _HTTP_OK:
+                    options = response.json()
+                    model_name = options.get("sd_model_checkpoint")
+                    return {
+                        "loaded": True,
+                        "model": model_name,
+                        "sd_server_url": sd_server_url,
+                    }
+        except httpx.HTTPError:
+            pass
 
-        if config is None:
         return {
             "loaded": False,
             "model": None,
-                "status": status.get("status"),
+            "sd_server_url": sd_server_url,
+            "error": "Cannot connect to sd-server",
         }
 
-        return {
-            "loaded": status.get("status") == "running",
-            "model": config.model,
-            "pid": status.get("pid"),
-            "port": config.port,
-            "status": status.get("status"),
-        }
-
-    @router.post("/switch")
-    async def switch_model(req: SwitchModelRequest) -> JSONResponse:
-        """Switch to a different model (hot reload)."""
-        model_path = Path(req.model)
-
-        # Validate model exists
-        if not model_path.exists():
-            raise HTTPException(status_code=400, detail=f"Model not found: {req.model}")
-
-        # Use existing reload logic
-        from tensors.server.models import ServerConfig  # noqa: PLC0415
-
-        new_config = ServerConfig(
-            model=req.model,
-            port=pm.config.port if pm.config else 1234,
-            args=pm.config.args if pm.config else [],
-        )
-
-        pm.stop()
-        pm.start(new_config)
-        ready = await pm.wait_ready()
-
-        if not ready:
-            return JSONResponse(
-                {"error": "sd-server failed to become ready", "model": req.model},
-                status_code=503,
-            )
-
-        return JSONResponse(
-            {
-                "ok": True,
-                "model": req.model,
-                "pid": pm.proc.pid if pm.proc else None,
-            }
-        )
-
     @router.get("/loras")
     def list_loras() -> dict[str, Any]:
         """List available LoRA files."""

tensors/server/process.py

@@ -1,92 +0,0 @@
-"""sd-server process lifecycle management."""
-
-from __future__ import annotations
-
-import asyncio
-import logging
-import os
-import shutil
-import signal
-import subprocess
-from typing import TYPE_CHECKING, Any
-
-import httpx
-
-if TYPE_CHECKING:
-    from tensors.server.models import ServerConfig
-
-logger = logging.getLogger(__name__)
-
-_HTTP_OK = 200
-
-SD_SERVER_BIN = shutil.which("sd-server") or "sd-server"
-
-
-class ProcessManager:
-    def __init__(self) -> None:
-        self.proc: subprocess.Popen[bytes] | None = None
-        self.config: ServerConfig | None = None
-
-    def build_cmd(self) -> list[str]:
-        if self.config is None:
-            raise RuntimeError("No config set")
-        cmd = [SD_SERVER_BIN, "-m", self.config.model, "--listen-port", str(self.config.port)]
-        cmd.extend(self.config.args)
-        return cmd
-
-    def start(self, config: ServerConfig) -> None:
-        if self.proc is not None and self.proc.poll() is None:
-            raise RuntimeError("Server already running — stop it first")
-        self.config = config
-        cmd = self.build_cmd()
-        # Inherit environment (important for HSA_OVERRIDE_GFX_VERSION on ROCm)
-        self.proc = subprocess.Popen(cmd, env=os.environ.copy())
-        logger.info("started sd-server pid=%d cmd=%s", self.proc.pid, cmd)
-
-    def stop(self) -> bool:
-        if self.proc is None or self.proc.poll() is not None:
-            self.proc = None
-            return False
-        self.proc.send_signal(signal.SIGTERM)
-        try:
-            self.proc.wait(timeout=10)
-        except subprocess.TimeoutExpired:
-            self.proc.kill()
-            self.proc.wait(timeout=5)
-        logger.info("stopped sd-server")
-        self.proc = None
-        return True
-
-    def status(self) -> dict[str, Any]:
-        if self.proc is None:
-            return {"running": False}
-        rc = self.proc.poll()
-        if rc is not None:
-            return {"running": False, "exit_code": rc}
-        return {
-            "running": True,
-            "pid": self.proc.pid,
-            "model": self.config.model if self.config else None,
-            "port": self.config.port if self.config else None,
-            "bind": f"http://127.0.0.1:{self.config.port}" if self.config else None,
-            "cmd": self.build_cmd(),
-        }
-
-    async def wait_ready(self, timeout: float = 120) -> bool:
-        """Poll sd-server root endpoint until it responds or timeout."""
-        if self.config is None:
-            return False
-        url = f"http://127.0.0.1:{self.config.port}/"
-        deadline = asyncio.get_event_loop().time() + timeout
-        async with httpx.AsyncClient() as client:
-            while asyncio.get_event_loop().time() < deadline:
-                if self.proc is not None and self.proc.poll() is not None:
-                    return False
-                try:
-                    r = await client.get(url, timeout=2)
-                    if r.status_code == _HTTP_OK:
-                        return True
-                except (httpx.ConnectError, httpx.TimeoutException, httpx.ReadError):
-                    pass
-                await asyncio.sleep(1)
-        return False

tensors/server/routes.py

@@ -3,53 +3,52 @@
 from __future__ import annotations
 
 import logging
-from typing import TYPE_CHECKING, Any
+from typing import Any
 
+import httpx
 from fastapi import APIRouter, Request, Response
 from fastapi.responses import JSONResponse, StreamingResponse
 
-from tensors.server.models import ReloadRequest, ServerConfig
-
-if TYPE_CHECKING:
-    from tensors.server.process import ProcessManager
-
 logger = logging.getLogger(__name__)
 
 
-def create_router(pm: ProcessManager) -> APIRouter:
+def create_router() -> APIRouter:
-    """Build a router with /status, /reload, and catch-all proxy."""
+    """Build a router with /status and catch-all proxy."""
     router = APIRouter()
 
     @router.get("/status")
-    def status() -> dict[str, Any]:
+    async def status(request: Request) -> dict[str, Any]:
-        return pm.status()
+        """Check if the external sd-server is reachable."""
-
+        sd_server_url = request.app.state.sd_server_url
-    @router.post("/reload")
+        try:
-    async def reload(req: ReloadRequest) -> Response:
+            async with httpx.AsyncClient(timeout=5) as client:
-        new_config = ServerConfig(
+                r = await client.get(sd_server_url)
-            model=req.model,
+                return {
-            port=pm.config.port if pm.config else 1234,
+                    "status": "ok",
-            args=pm.config.args if pm.config else [],
+                    "sd_server_url": sd_server_url,
-        )
+                    "sd_server_status": r.status_code,
-        pm.stop()
+                }
-        pm.start(new_config)
+        except httpx.HTTPError as e:
-        ready = await pm.wait_ready()
+            return {
-        if not ready:
+                "status": "error",
-            return JSONResponse({"error": "sd-server failed to become ready", "model": req.model}, status_code=503)
+                "sd_server_url": sd_server_url,
-        return JSONResponse({"ok": True, "model": req.model, "pid": pm.proc.pid if pm.proc else None})
+                "error": str(e),
+            }
 
     @router.api_route("/{path:path}", methods=["GET", "POST", "PUT", "DELETE", "PATCH", "HEAD", "OPTIONS"])
     async def proxy(request: Request, path: str) -> Response:
-        if pm.proc is None or pm.proc.poll() is not None:
+        """Proxy all requests to the external sd-server."""
-            return JSONResponse({"error": "sd-server is not running"}, status_code=503)
+        sd_server_url = request.app.state.sd_server_url
-        assert pm.config is not None
+        url = f"{sd_server_url}/{path}"
-        url = f"http://127.0.0.1:{pm.config.port}/{path}"
         if request.url.query:
             url = f"{url}?{request.url.query}"
+
         body = await request.body()
         headers = dict(request.headers)
         headers.pop("host", None)
         client = request.app.state.client
+
+        try:
             upstream = await client.request(
                 method=request.method,
                 url=url,
@@ -62,5 +61,15 @@ def create_router(pm: ProcessManager) -> APIRouter:
                 status_code=upstream.status_code,
                 headers=dict(upstream.headers),
             )
+        except httpx.ConnectError:
+            return JSONResponse(
+                {"error": f"Cannot connect to sd-server at {sd_server_url}"},
+                status_code=503,
+            )
+        except httpx.TimeoutException:
+            return JSONResponse(
+                {"error": f"Timeout connecting to sd-server at {sd_server_url}"},
+                status_code=504,
+            )
 
     return router

tests/test_server.py

@@ -2,110 +2,51 @@
 
 from __future__ import annotations
 
-from unittest.mock import AsyncMock, MagicMock, patch
+from unittest.mock import AsyncMock
 
 import httpx
 import pytest
+import respx
 from fastapi.testclient import TestClient
 
 from tensors.server import create_app
-from tensors.server.models import ServerConfig
-from tensors.server.process import ProcessManager
-
-
-@pytest.fixture()
-def pm() -> ProcessManager:
-    return ProcessManager()
 
 
 @pytest.fixture()
 def api() -> TestClient:
-    return TestClient(create_app())
+    """Create test client with mock sd-server URL."""
-
+    return TestClient(create_app(sd_server_url="http://mock-sd-server:1234"))
-
-def _get_pm(api: TestClient) -> ProcessManager:
-    return api.app.state.pm  # type: ignore[no-any-return, attr-defined]
 
 
 class TestStatus:
-    def test_not_running(self, api: TestClient) -> None:
+    @respx.mock
-        r = api.get("/status")
+    def test_status_when_backend_reachable(self) -> None:
-        assert r.status_code == 200
+        """Test status endpoint when sd-server is reachable."""
-        assert r.json()["running"] is False
+        respx.get("http://mock-sd-server:1234/").mock(return_value=httpx.Response(200))
 
-    def test_running(self, api: TestClient) -> None:
+        with TestClient(create_app(sd_server_url="http://mock-sd-server:1234")) as client:
-        pm = _get_pm(api)
+            r = client.get("/status")
-        mock_proc = MagicMock()
-        mock_proc.poll.return_value = None
-        mock_proc.pid = 999
-        pm.proc = mock_proc
-        pm.config = ServerConfig(model="/m.safetensors")
-        r = api.get("/status")
-        data = r.json()
-        assert data["running"] is True
-        assert data["pid"] == 999
-
-    def test_exited(self, api: TestClient) -> None:
-        pm = _get_pm(api)
-        mock_proc = MagicMock()
-        mock_proc.poll.return_value = 1
-        pm.proc = mock_proc
-        r = api.get("/status")
-        data = r.json()
-        assert data["running"] is False
-        assert data["exit_code"] == 1
-
-
-class TestReload:
-    @patch.object(ProcessManager, "wait_ready", new_callable=AsyncMock, return_value=True)
-    @patch("tensors.server.process.subprocess.Popen")
-    def test_reload_swaps_model(self, mock_popen: MagicMock, mock_ready: AsyncMock, api: TestClient) -> None:
-        pm = _get_pm(api)
-        pm.config = ServerConfig(model="/old.gguf", port=5555, args=["--fa"])
-        mock_popen.return_value.pid = 42
-        mock_popen.return_value.poll.return_value = None
-        r = api.post("/reload", json={"model": "/new.gguf"})
             assert r.status_code == 200
             data = r.json()
-        assert data["ok"] is True
+            assert data["status"] == "ok"
-        assert data["model"] == "/new.gguf"
+            assert data["sd_server_url"] == "http://mock-sd-server:1234"
-        assert data["pid"] == 42
-        # Verify new config preserved port and args from previous config
-        assert pm.config is not None
-        assert pm.config.port == 5555
-        assert pm.config.args == ["--fa"]
-        assert pm.config.model == "/new.gguf"
 
-    @patch.object(ProcessManager, "wait_ready", new_callable=AsyncMock, return_value=False)
+    @respx.mock
-    @patch("tensors.server.process.subprocess.Popen")
+    def test_status_when_backend_unreachable(self) -> None:
-    def test_reload_fails_when_not_ready(self, mock_popen: MagicMock, mock_ready: AsyncMock, api: TestClient) -> None:
+        """Test status endpoint when sd-server is not reachable."""
-        pm = _get_pm(api)
+        respx.get("http://mock-sd-server:1234/").mock(side_effect=httpx.ConnectError("Connection refused"))
-        pm.config = ServerConfig(model="/old.gguf")
-        mock_popen.return_value.pid = 43
-        mock_popen.return_value.poll.return_value = None
-        r = api.post("/reload", json={"model": "/bad.gguf"})
-        assert r.status_code == 503
-        assert "failed" in r.json()["error"]
 
-    def test_reload_requires_model(self, api: TestClient) -> None:
+        with TestClient(create_app(sd_server_url="http://mock-sd-server:1234")) as client:
-        r = api.post("/reload", json={})
+            r = client.get("/status")
-        assert r.status_code == 422
+            assert r.status_code == 200
+            data = r.json()
+            assert data["status"] == "error"
+            assert "Connection refused" in data["error"]
 
 
 class TestProxy:
-    def test_proxy_503_when_not_running(self, api: TestClient) -> None:
-        r = api.get("/v1/models")
-        assert r.status_code == 503
-        assert "not running" in r.json()["error"]
-
     def test_proxy_forwards_request(self, api: TestClient) -> None:
-        pm = _get_pm(api)
+        """Test proxy forwards GET requests to backend."""
-        mock_proc = MagicMock()
-        mock_proc.poll.return_value = None
-        mock_proc.pid = 100
-        pm.proc = mock_proc
-        pm.config = ServerConfig(model="/m.gguf", port=1234)
-
         upstream_response = httpx.Response(
             200,
             json={"data": [{"id": "model-1"}]},
@@ -114,6 +55,7 @@ class TestProxy:
         mock_client = AsyncMock()
         mock_client.request.return_value = upstream_response
         api.app.state.client = mock_client  # type: ignore[attr-defined]
+        api.app.state.sd_server_url = "http://mock-sd-server:1234"  # type: ignore[attr-defined]
 
         r = api.get("/v1/models")
         assert r.status_code == 200
@@ -121,52 +63,38 @@ class TestProxy:
         mock_client.request.assert_called_once()
 
     def test_proxy_forwards_post_with_body(self, api: TestClient) -> None:
-        pm = _get_pm(api)
+        """Test proxy forwards POST requests with body."""
-        mock_proc = MagicMock()
-        mock_proc.poll.return_value = None
-        mock_proc.pid = 100
-        pm.proc = mock_proc
-        pm.config = ServerConfig(model="/m.gguf", port=1234)
-
         upstream_response = httpx.Response(200, json={"ok": True})
         mock_client = AsyncMock()
         mock_client.request.return_value = upstream_response
         api.app.state.client = mock_client  # type: ignore[attr-defined]
+        api.app.state.sd_server_url = "http://mock-sd-server:1234"  # type: ignore[attr-defined]
 
-        r = api.post("/v1/chat/completions", json={"prompt": "hello"})
+        r = api.post("/sdapi/v1/txt2img", json={"prompt": "hello"})
         assert r.status_code == 200
         mock_client.request.assert_called_once()
 
+    def test_proxy_503_on_connect_error(self, api: TestClient) -> None:
+        """Test proxy returns 503 when backend is unreachable."""
+        mock_client = AsyncMock()
+        mock_client.request.side_effect = httpx.ConnectError("Connection refused")
+        api.app.state.client = mock_client  # type: ignore[attr-defined]
+        api.app.state.sd_server_url = "http://mock-sd-server:1234"  # type: ignore[attr-defined]
 
-class TestProcessManager:
+        r = api.get("/v1/models")
-    def test_status_not_running(self, pm: ProcessManager) -> None:
+        assert r.status_code == 503
-        assert pm.status() == {"running": False}
+        assert "Cannot connect" in r.json()["error"]
 
-    def test_build_cmd(self, pm: ProcessManager) -> None:
+    def test_proxy_504_on_timeout(self, api: TestClient) -> None:
-        pm.config = ServerConfig(model="/m.gguf", port=1234, args=["--fa"])
+        """Test proxy returns 504 on timeout."""
-        cmd = pm.build_cmd()
+        mock_client = AsyncMock()
-        assert "/m.gguf" in cmd
+        mock_client.request.side_effect = httpx.TimeoutException("Timeout")
-        assert "--fa" in cmd
+        api.app.state.client = mock_client  # type: ignore[attr-defined]
-        assert "1234" in cmd
+        api.app.state.sd_server_url = "http://mock-sd-server:1234"  # type: ignore[attr-defined]
 
-    def test_build_cmd_no_config(self, pm: ProcessManager) -> None:
+        r = api.get("/v1/models")
-        with pytest.raises(RuntimeError, match="No config"):
+        assert r.status_code == 504
-            pm.build_cmd()
+        assert "Timeout" in r.json()["error"]
-
-    @patch("tensors.server.process.subprocess.Popen")
-    def test_start_and_stop(self, mock_popen: MagicMock, pm: ProcessManager) -> None:
-        mock_popen.return_value.pid = 77
-        mock_popen.return_value.poll.return_value = None
-        mock_popen.return_value.wait.return_value = 0
-        pm.start(ServerConfig(model="/m.gguf"))
-        assert pm.proc is not None
-        assert pm.stop() is True
-        assert pm.proc is None
-
-    def test_server_config_defaults(self) -> None:
-        cfg = ServerConfig(model="/m.gguf")
-        assert cfg.port == 1234
-        assert cfg.args == []
 
 
 # =============================================================================

uv.lock

@@ -707,7 +707,7 @@ wheels = [
 
 [[package]]
 name = "tensors"
-version = "0.1.9"
+version = "0.1.12"
 source = { editable = "." }
 dependencies = [
     { name = "httpx" },