openenv_version: "1.0"

environment:
  name: smart-study-session-optimizer
  description: >
    An RL environment where an agent learns to balance studying and breaks
    to maximize productivity without burning out. Inspired by my own terrible
    study habits during college.
  version: "1.0.0"
  author: "Hackathon Participant"

observation_space:
  type: object
  properties:
    energy:
      type: number
      minimum: 0.0
      maximum: 100.0
      description: Current energy level of the student (0=exhausted, 100=fully rested)
    focus:
      type: number
      minimum: 0.0
      maximum: 100.0
      description: Focus/concentration score right now
    time_studied:
      type: number
      minimum: 0.0
      description: Total minutes studied in this session so far
    session_minutes:
      type: number
      minimum: 0.0
      description: Total elapsed minutes in the current session
    consecutive_study_minutes:
      type: number
      minimum: 0.0
      description: How many minutes studied in a row without a break
    break_count:
      type: integer
      minimum: 0
      description: Number of breaks taken so far
    productive_minutes:
      type: number
      minimum: 0.0
      description: Minutes where focus was above 60 (counted as productive)

action_space:
  type: integer
  values:
    - value: 0
      name: study
      description: Continue studying for the next time step (5 minutes)
    - value: 1
      name: short_break
      description: Take a short 10-minute break to recover some energy
    - value: 2
      name: long_break
      description: Take a long 20-minute break for full energy recovery

endpoints:
  reset:
    method: POST
    path: /reset
    description: Reset the environment to initial state, returns first observation
  step:
    method: POST
    path: /step
    description: Execute one action in the environment
    request_body:
      action:
        type: integer
        description: Action to take (0=study, 1=short break, 2=long break)
    response_body:
      observation:
        type: object
        description: New observation after action
      reward:
        type: number
        description: Reward signal for this step
      done:
        type: boolean
        description: Whether the episode is finished
      info:
        type: object
        description: Extra info for debugging/grading
  state:
    method: GET
    path: /state
    description: Get current environment state without modifying it
  health:
    method: GET
    path: /health
    description: Health check endpoint

tasks:
  - id: survive_8_hours
    name: "Survive 8 Hours"
    description: >
      Study for 8 hours (480 minutes) without energy hitting zero.
      Score based on remaining energy (clamped 0.1 to 0.9).
    max_steps: 120
    scoring: 0.1 + (energy / 100) * 0.8
    success_threshold: 0.5

  - id: maximize_productivity
    name: "Maximize Productivity"
    description: >
      Get as many productive minutes (focus > 60) as possible.
      Score based on focus and energy (clamped 0.1 to 0.9).
    max_steps: 60
    scoring: 0.1 + (focus / 200) * 0.8 + (energy / 200) * 0.8
    success_threshold: 0.5

  - id: smart_break_taker
    name: "Smart Break Taker"
    description: >
      Only take breaks when energy is low (below 40%).
      Score based on energy and focus (clamped 0.1 to 0.9).
    max_steps: 80
    scoring: 0.1 + (energy / 200) * 0.8 + (focus / 200) * 0.8
    success_threshold: 0.6