Email Triage OpenEnv 📧

An OpenEnv environment where an AI agent learns to triage real-world emails — classify spam, categorize messages, and perform full professional triage.

Environment Description

Email triage is something every professional does daily. This environment simulates that task at three levels of difficulty, giving AI agents meaningful, graded feedback at every step.

Tasks

Task	Difficulty	Description	Reward
spam_detection	Easy	Classify emails as spam or ham	1.0 correct / 0.0 wrong
categorization	Medium	Assign to work / personal / shopping / promotions	1.0 correct / 0.0 wrong
full_triage	Hard	Set priority + write summary + draft reply	Partial (0.0–1.0)

Observation Space

{
  "email_id": "string",
  "subject": "string",
  "sender": "string",
  "body": "string",
  "task_type": "string",
  "task_description": "string",
  "available_actions": ["list", "of", "valid", "labels"],
  "feedback": "string (after step)",
  "done": false,
  "reward": 0.0
}

Action Space

{
  "action_type": "classify | categorize | prioritize",
  "label": "the predicted label",
  "summary": "one sentence summary (hard task only)",
  "reply_draft": "short reply draft (hard task only)"
}

Reward Function

• Easy/Medium: Binary — 1.0 for correct, 0.0 for wrong
• Hard: Partial scoring
  • Priority accuracy: up to 0.4 (partial credit for close guesses)
  • Summary quality: up to 0.3 (keyword overlap with expected)
  • Reply quality: up to 0.3 (key action words present)

Setup & Usage

Local (Python)

pip install -r requirements.txt
uvicorn server.app:app --host 0.0.0.0 --port 7860

Docker

docker build -t email-triage-env .
docker run -p 7860:7860 email-triage-env

Run Inference

export HF_TOKEN=your_token_here
export ENV_BASE_URL=http://localhost:7860
python inference.py

API Endpoints

Endpoint	Method	Description
/health	GET	Health check
/reset	POST	Start new episode
/step	POST	Take an action
/state	GET	Get current state
/tasks	GET	List all tasks

Baseline Scores

Task	Score
spam_detection	~0.85
categorization	~0.75
full_triage	~0.55
Overall	~0.72

Environment Variables

Variable	Description
API_BASE_URL	LLM API endpoint
MODEL_NAME	Model identifier
HF_TOKEN	Hugging Face API token
ENV_BASE_URL	Environment server URL