Prepare HF deploy with OpenAI token

.env.example

@@ -1,8 +1,16 @@
+# Default profile: Hugging Face Router (OpenAI-compatible API)
 API_BASE_URL=https://router.huggingface.co/v1
 MODEL_NAME=Qwen/Qwen2.5-72B-Instruct
-OPENAI_API_KEY=<your_openai_compatible_api_key>
+OPENAI_API_KEY=<your_openai_or_router_api_key>
+
+# Optional backward-compatible fallback. Used only when OPENAI_API_KEY is unset.
 HF_TOKEN=<optional_fallback_token>
 
+# OpenAI direct profile (uncomment for OpenAI access token usage):
+# API_BASE_URL=https://api.openai.com/v1
+# MODEL_NAME=gpt-4o-mini
+# OPENAI_API_KEY=<your_openai_access_token>
+
 # Optional runtime knobs
 LOCAL_IMAGE_NAME=
 MY_ENV_V4_TASK=easy-command-typo

Dockerfile

@@ -10,4 +10,4 @@ COPY . .
 ENV PORT=7860
 EXPOSE 7860
 
-CMD ["python", "-m", "uvicorn", "server.app:app", "--host", "0.0.0.0", "--port", "7860"]
+CMD ["python", "-m", "server.app"]

README.md

@@ -220,6 +220,14 @@ export HF_TOKEN="<your_token>"
 export LOCAL_IMAGE_NAME="<local_env_image_name>"
 ```
 
+If you want to use an OpenAI access token directly:
+
+```bash
+export API_BASE_URL="https://api.openai.com/v1"
+export MODEL_NAME="gpt-4o-mini"
+export OPENAI_API_KEY="<your_openai_access_token>"
+```
+
 ## 8. Run Inference
 
 Offline/local mode:
@@ -330,6 +338,43 @@ Server endpoints used by validators:
 - `GET /state`
 - `GET /health`
 
-## 12. One-line Presentation Summary
+## 12. Deploy to Hugging Face Space (OpenAI Token)
+
+This repository is already configured for Docker Spaces (`sdk: docker` in this README front matter).
+
+1. Create a new Hugging Face Space with SDK set to `Docker`.
+2. Push this repository to the Space git remote.
+3. In Space Settings -> Variables and secrets, add these Secrets:
+
+```text
+OPENAI_API_KEY=<your_openai_access_token>
+API_BASE_URL=https://api.openai.com/v1
+MODEL_NAME=gpt-4o-mini
+```
+
+4. Optional Secrets:
+
+```text
+HF_TOKEN=<optional_fallback_token>
+OFFLINE_INFERENCE=0
+MAX_STEPS=8
+TEMPERATURE=0.2
+MAX_TOKENS=120
+```
+
+5. Keep the app port as `7860` (already configured).
+6. Wait for build completion, then verify:
+
+```bash
+curl -sS https://<your-space-name>.hf.space/health
+curl -sS -X POST https://<your-space-name>.hf.space/reset -H 'Content-Type: application/json' -d '{}'
+```
+
+Notes:
+
+- `.env.example` is for local development reference only. Hugging Face Spaces use Secrets/Variables from Space Settings.
+- Runtime code reads `OPENAI_API_KEY` first and falls back to `HF_TOKEN` when `OPENAI_API_KEY` is not provided.
+
+## 13. One-line Presentation Summary
 
 We built an OpenEnv-compliant reinforcement learning environment where AI agents learn to debug real CI/CD pipelines using multi-step reasoning, hybrid grading, anti-hacking safeguards, and robust reward shaping.