refactor: update Dockerfile and pyproject.toml for dependency management; remove FastAPI server implementation

Dockerfile

@@ -20,7 +20,7 @@ RUN apt-get update && apt-get install -y supervisor && rm -rf /var/lib/apt/lists
 # Install Python dependencies
 COPY python/requirements.txt ./requirements.txt
 RUN pip install --no-cache-dir -r requirements.txt || \
-    pip install --no-cache-dir openai fastapi uvicorn httpx pydantic "pydantic>=2.0.0" requests
+    pip install --no-cache-dir openai httpx pydantic "pydantic>=2.0.0" requests
 
 # Copy Go binary
 COPY --from=builder /app/gridmind-server /usr/local/bin/gridmind-server

pyproject.toml

@@ -54,8 +54,7 @@ Repository = "https://github.com/meta-pytorch/OpenEnv"
 Documentation = "https://github.com/meta-pytorch/OpenEnv"
 
 [project.scripts]
-server = "server.app:main"
-gridmind-server = "server.app:main"
+gridmind-eval = "python.inference:main"
 
 [tool.setuptools]
 packages = ["python", "server"]

server/app.py

@@ -1,84 +0,0 @@
-"""
-GridMind-RL FastAPI Server Wrapper
-Proxies requests to the Go environment server (or provides fallback endpoints)
-"""
-
-from fastapi import FastAPI
-from fastapi.responses import JSONResponse
-import os
-import httpx
-
-app = FastAPI(title="GridMind-RL", version="1.0.0")
-
-# Go server address (set via environment or default)
-GO_SERVER_URL = os.getenv("GO_SERVER_URL", "http://localhost:8000")
-
-# Timeout for Go server calls
-TIMEOUT = 30
-
-
-@app.get("/health")
-async def health():
-    """Health check endpoint."""
-    try:
-        async with httpx.AsyncClient(timeout=TIMEOUT) as client:
-            resp = await client.get(f"{GO_SERVER_URL}/health")
-            return resp.json()
-    except Exception:
-        # Fallback response if Go server unreachable
-        return {"status": "ok", "mode": "python"}
-
-
-@app.get("/state")
-async def get_state():
-    """Get environment state."""
-    try:
-        async with httpx.AsyncClient(timeout=TIMEOUT) as client:
-            resp = await client.get(f"{GO_SERVER_URL}/state")
-            return resp.json()
-    except Exception as e:
-        return JSONResponse({"error": str(e)}, status_code=503)
-
-
-@app.post("/reset")
-async def reset(request: dict):
-    """Reset environment."""
-    try:
-        async with httpx.AsyncClient(timeout=TIMEOUT) as client:
-            resp = await client.post(f"{GO_SERVER_URL}/reset", json=request)
-            return resp.json()
-    except Exception as e:
-        return JSONResponse({"error": str(e)}, status_code=503)
-
-
-@app.post("/step")
-async def step(request: dict):
-    """Step environment."""
-    try:
-        async with httpx.AsyncClient(timeout=TIMEOUT) as client:
-            resp = await client.post(f"{GO_SERVER_URL}/step", json=request)
-            return resp.json()
-    except Exception as e:
-        return JSONResponse({"error": str(e)}, status_code=503)
-
-
-@app.get("/grade")
-async def grade():
-    """Grade environment."""
-    try:
-        async with httpx.AsyncClient(timeout=TIMEOUT) as client:
-            resp = await client.get(f"{GO_SERVER_URL}/grade")
-            return resp.json()
-    except Exception as e:
-        return JSONResponse({"error": str(e)}, status_code=503)
-
-
-def main():
-    """Entry point for server."""
-    import uvicorn
-    port = int(os.getenv("PORT", 7860))
-    uvicorn.run(app, host="0.0.0.0", port=port)
-
-
-if __name__ == "__main__":
-    main()

test_suite_results.json

@@ -0,0 +1,307 @@
+{
+  "passed": 50,
+  "failed": 0,
+  "warnings": 0,
+  "details": [
+    {
+      "id": "B1",
+      "description": "API_BASE_URL has default",
+      "status": "PASS",
+      "details": "https://openrouter.ai/api/v1"
+    },
+    {
+      "id": "B2",
+      "description": "MODEL_NAME has default",
+      "status": "PASS",
+      "details": "meta-llama/llama-3.3-70b-instruct:free"
+    },
+    {
+      "id": "B3",
+      "description": "HF_TOKEN is mandatory (raises ValueError)",
+      "status": "PASS",
+      "details": "checked in python/inference.py line 57-58"
+    },
+    {
+      "id": "B4",
+      "description": "OpenAI client initialized",
+      "status": "PASS",
+      "details": "base_url and api_key from env vars"
+    },
+    {
+      "id": "B5",
+      "description": "HF_TOKEN is PRIMARY key",
+      "status": "PASS",
+      "details": "OPENAI_API_KEY is fallback only"
+    },
+    {
+      "id": "C1",
+      "description": "[START] format exists",
+      "status": "PASS",
+      "details": ""
+    },
+    {
+      "id": "C2",
+      "description": "[STEP] format exists",
+      "status": "PASS",
+      "details": ""
+    },
+    {
+      "id": "C3",
+      "description": "[END] format exists",
+      "status": "PASS",
+      "details": ""
+    },
+    {
+      "id": "C4",
+      "description": "Reward formatted to 2 decimal places",
+      "status": "PASS",
+      "details": ""
+    },
+    {
+      "id": "C5",
+      "description": "Lowercase booleans (true/false)",
+      "status": "PASS",
+      "details": ""
+    },
+    {
+      "id": "C6",
+      "description": "error field uses null",
+      "status": "PASS",
+      "details": ""
+    },
+    {
+      "id": "D1",
+      "description": "OpenEnv spec structure valid, name=gridmind-rl",
+      "status": "PASS",
+      "details": ""
+    },
+    {
+      "id": "D1b",
+      "description": "Port is 7860: 7860",
+      "status": "PASS",
+      "details": ""
+    },
+    {
+      "id": "D1c",
+      "description": "Has 3 tasks: 3",
+      "status": "PASS",
+      "details": ""
+    },
+    {
+      "id": "D2",
+      "description": "All OpenEnv endpoints declared: 6/6",
+      "status": "PASS",
+      "details": ""
+    },
+    {
+      "id": "E1",
+      "description": "Exactly 3 tasks: 3",
+      "status": "PASS",
+      "details": ""
+    },
+    {
+      "id": "E1.1",
+      "description": "Task 1 difficulty is easy",
+      "status": "PASS",
+      "details": "expected easy"
+    },
+    {
+      "id": "E1.2",
+      "description": "Task 2 difficulty is medium",
+      "status": "PASS",
+      "details": "expected medium"
+    },
+    {
+      "id": "E1.3",
+      "description": "Task 3 difficulty is hard",
+      "status": "PASS",
+      "details": "expected hard"
+    },
+    {
+      "id": "E2",
+      "description": "Task 1 grader exists",
+      "status": "PASS",
+      "details": ""
+    },
+    {
+      "id": "E2",
+      "description": "Task 2 grader exists",
+      "status": "PASS",
+      "details": ""
+    },
+    {
+      "id": "E2",
+      "description": "Task 3 grader exists",
+      "status": "PASS",
+      "details": ""
+    },
+    {
+      "id": "E5",
+      "description": "Exploit detection exists",
+      "status": "PASS",
+      "details": ""
+    },
+    {
+      "id": "E6.1",
+      "description": "Task 1 weights sum: 1.00",
+      "status": "PASS",
+      "details": ""
+    },
+    {
+      "id": "E6.2",
+      "description": "Task 2 weights sum: 1.00",
+      "status": "PASS",
+      "details": ""
+    },
+    {
+      "id": "E6.3",
+      "description": "Task 3 weights sum: 1.00",
+      "status": "PASS",
+      "details": ""
+    },
+    {
+      "id": "F1",
+      "description": "All 7 reward components exist: 7/7",
+      "status": "PASS",
+      "details": ""
+    },
+    {
+      "id": "F2",
+      "description": "Reward computed every step",
+      "status": "PASS",
+      "details": ""
+    },
+    {
+      "id": "F3",
+      "description": "Penalties for bad behaviors",
+      "status": "PASS",
+      "details": ""
+    },
+    {
+      "id": "F4",
+      "description": "Reward aggregated properly",
+      "status": "PASS",
+      "details": ""
+    },
+    {
+      "id": "G1",
+      "description": "Multi-stage build (Go builder + Python runtime)",
+      "status": "PASS",
+      "details": ""
+    },
+    {
+      "id": "G2",
+      "description": "Go server compiled",
+      "status": "PASS",
+      "details": ""
+    },
+    {
+      "id": "G3",
+      "description": "supervisord manages processes",
+      "status": "PASS",
+      "details": ""
+    },
+    {
+      "id": "G4",
+      "description": "Go server on port 7860",
+      "status": "PASS",
+      "details": ""
+    },
+    {
+      "id": "G5",
+      "description": "Dashboard on port 7861",
+      "status": "PASS",
+      "details": ""
+    },
+    {
+      "id": "G7",
+      "description": "Both ports exposed",
+      "status": "PASS",
+      "details": ""
+    },
+    {
+      "id": "G8",
+      "description": "Dockerfile syntax valid",
+      "status": "PASS",
+      "details": ""
+    },
+    {
+      "id": "H1",
+      "description": "README has Overview/Motivation",
+      "status": "PASS",
+      "details": ""
+    },
+    {
+      "id": "H2",
+      "description": "README documents Observation Space",
+      "status": "PASS",
+      "details": ""
+    },
+    {
+      "id": "H3",
+      "description": "README documents Action Space",
+      "status": "PASS",
+      "details": ""
+    },
+    {
+      "id": "H4",
+      "description": "README has task descriptions",
+      "status": "PASS",
+      "details": ""
+    },
+    {
+      "id": "H5",
+      "description": "README has setup/usage",
+      "status": "PASS",
+      "details": ""
+    },
+    {
+      "id": "H6",
+      "description": "README mentions baseline",
+      "status": "PASS",
+      "details": ""
+    },
+    {
+      "id": "H7",
+      "description": "README mentions OpenEnv/HF",
+      "status": "PASS",
+      "details": ""
+    },
+    {
+      "id": "I1",
+      "description": "All 3 task scores present: 3",
+      "status": "PASS",
+      "details": ""
+    },
+    {
+      "id": "I2",
+      "description": "All scores in [0.0, 1.0]",
+      "status": "PASS",
+      "details": ""
+    },
+    {
+      "id": "I3",
+      "description": "Has model and api_base fields",
+      "status": "PASS",
+      "details": ""
+    },
+    {
+      "id": "J1",
+      "description": "requirements.txt exists and includes required packages",
+      "status": "PASS",
+      "details": ""
+    },
+    {
+      "id": "J2",
+      "description": "openai package is importable",
+      "status": "PASS",
+      "details": ""
+    },
+    {
+      "id": "J4",
+      "description": "requests package is importable",
+      "status": "PASS",
+      "details": ""
+    }
+  ]
+}