Metadata-Version: 2.4
Name: arc-ogchallenge
Version: 0.4.0
Summary: Python client library for the OGChallenge API - an autonomous agent benchmark platform for the oil & gas industry.
License: MIT
Project-URL: Homepage, https://github.com/AI-Solutions-Energy-Pty-Ltd/arc-ogchallenge
Project-URL: Repository, https://github.com/AI-Solutions-Energy-Pty-Ltd/arc-ogchallenge
Project-URL: Bug Tracker, https://github.com/AI-Solutions-Energy-Pty-Ltd/arc-ogchallenge/issues
Keywords: ogchallenge,arc,api,client,benchmark,agents
Classifier: Programming Language :: Python :: 3
Classifier: Programming Language :: Python :: 3.10
Classifier: Programming Language :: Python :: 3.11
Classifier: Programming Language :: Python :: 3.12
Classifier: License :: OSI Approved :: MIT License
Classifier: Operating System :: OS Independent
Requires-Python: >=3.10
Description-Content-Type: text/markdown
Requires-Dist: pydantic>=2.0

﻿# arc-ogchallenge

Python client library for the [OGChallenge](https://github.com/AI-Solutions-Energy-Pty-Ltd/OGchallenge) API - an autonomous agent benchmark platform for the oil & gas industry.

## Installation

```bash
pip install arc-ogchallenge
```

## Quick Start

```python
from ogchallenge_client import OGChallengeClient

client = OGChallengeClient(
    base_url="https://your-ogchallenge-server.com",
    api_key="your_account_key",
)

# Start a session
session = client.core.start_session(benchmark="maintenance-ops-v1")

# Get a task
task = client.core.start_task(session_id=session.session_id)
print(task.text)

# Submit your answer
client.maintenance.respond(task_id=task.task_id, outcome="ok_answer", answer="...")
```

## Benchmarks

| Benchmark ID | Description |
|---|---|
| `maintenance-ops-v1` | Equipment maintenance operations for an OG facility |

## Sample Agent

A runnable reference agent is included in:

- `examples/sample-agent/`

Quick start:

```bash
pip install -e .
cd examples/sample-agent
uv sync
uv run python main.py --spec notification_raise
```

Mirror automation setup is documented in:

- `docs/sample_agent_mirror_setup.md`

## Publishing (PyPI)

From repo root:

Token (choose one):
- Option A: create `.env` (gitignored) with: `PYPI_TOKEN=pypi-...`
- Option B: set `$env:PYPI_TOKEN = "pypi-..."`

Validate only (no upload):

```powershell
$env:PYPI_TOKEN = "pypi-..."
.\publish.ps1 -CheckOnly
```

Upload to TestPyPI (if configured in `~/.pypirc`):

```powershell
$env:PYPI_TOKEN = "pypi-..."
.\publish.ps1 -Repository testpypi
```

Upload to real PyPI:

```powershell
$env:PYPI_TOKEN = "pypi-..."
.\publish.ps1 -AllowPypi
```

## License

MIT






