Spaces:
Sleeping
Sleeping
File size: 4,068 Bytes
11fd923 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 | """
Bug Report Structuring Environment - Data Models
Defines typed request/response models for the OpenEnv API:
- BugReportAction: What the agent submits (structured bug report)
- BugReportObservation: What the environment returns (feedback + score)
- BugReportState: Episode metadata
"""
from pydantic import BaseModel, Field
from typing import Optional, Dict, List
# ─── Request Models ───────────────────────────────────────────────
class ResetRequest(BaseModel):
"""Request body for POST /reset"""
task_id: Optional[str] = Field(
default=None,
description="Task difficulty: 'easy', 'medium', or 'hard'. Random if not specified."
)
seed: Optional[int] = Field(default=None, description="Random seed for reproducibility")
episode_id: Optional[str] = Field(default=None, description="Custom episode ID")
class BugReportAction(BaseModel):
"""
The structured bug report submitted by the agent.
This is the action the agent takes at each step.
"""
title: str = Field(
...,
description="Clear, concise bug title"
)
steps_to_reproduce: str = Field(
...,
description="Step-by-step reproduction instructions"
)
expected_behavior: str = Field(
...,
description="What should happen"
)
actual_behavior: str = Field(
...,
description="What actually happens"
)
severity: str = Field(
...,
description="Severity level: 'low', 'medium', 'high', or 'critical'"
)
environment: str = Field(
...,
description="OS, browser, version, platform info"
)
additional_notes: Optional[str] = Field(
default="",
description="Any additional relevant information"
)
class StepRequest(BaseModel):
"""Request body for POST /step"""
action: BugReportAction
# ─── Response Models ──────────────────────────────────────────────
class BugReportObservation(BaseModel):
"""
What the environment returns after reset() or step().
Contains the raw report, feedback, and scoring.
"""
raw_report: str = Field(
...,
description="The messy, unstructured bug report to process"
)
feedback: str = Field(
default="",
description="Grading feedback explaining the score"
)
score: float = Field(
default=0.0,
description="Overall score from 0.0 to 1.0"
)
field_scores: Dict[str, float] = Field(
default_factory=dict,
description="Per-field scores: title, steps, expected, actual, severity, environment"
)
done: bool = Field(
default=False,
description="Whether the episode is complete"
)
reward: float = Field(
default=0.0,
description="Reward signal for this step"
)
step_count: int = Field(
default=0,
description="Current step number"
)
task_id: str = Field(
default="",
description="Current task identifier"
)
max_steps: int = Field(
default=5,
description="Maximum steps allowed for this task"
)
class BugReportState(BaseModel):
"""
Episode state metadata returned by GET /state.
"""
episode_id: str = Field(default="", description="Unique episode identifier")
step_count: int = Field(default=0, description="Steps taken so far")
task_id: str = Field(default="", description="Current task ID")
max_steps: int = Field(default=5, description="Maximum steps for this task")
current_score: float = Field(default=0.0, description="Best score achieved so far")
best_score: float = Field(default=0.0, description="Best score across all steps")
done: bool = Field(default=False, description="Whether episode is complete")
rewards: List[float] = Field(default_factory=list, description="Rewards per step")
|