README.md

---
title: Pr Webhook
emoji: 📊
colorFrom: blue
colorTo: gray
sdk: docker
pinned: false
---

# PR Webhook with Code Review Agent

A webhook service that integrates with a code review agent powered by Pydantic AI and enhanced with Logfire tracing.

## Features

- **Code Review Agent**: Automated code review using Pydantic AI
- **Logfire Tracing**: Comprehensive tracing and monitoring of agent runs
- **Webhook Integration**: Handles pull request webhooks
- **Structured Output**: Generates detailed code review reports

## Setup

### Environment Variables

Create a `.env` file with the following variables:

```bash
# Logfire Configuration
LOGFIRE_API_KEY=your_logfire_api_key_here

# Code Review Model
CODE_REVIEW_MODEL=google-gla:gemini-2.5-pro
```

### Installation

```bash
pip install -r requirements.txt
```

## Usage

### Running the Agent

```bash
python agent/code_review.py
```

### Using the Webhook

The service will automatically trace all agent runs with detailed metrics including:
- Execution time
- Input/output sizes
- Error tracking
- Performance metrics

## Tracing Features

The agent now includes comprehensive tracing with Logfire:

- **Span Tracking**: Each major operation is wrapped in a tracing span
- **Event Logging**: Key milestones are logged as events
- **Error Handling**: Exceptions are automatically captured and traced
- **Performance Metrics**: Input/output sizes and execution times are tracked
- **Context Attributes**: Relevant metadata is attached to each span

## Architecture

- `agent/code_review.py`: Main code review agent with Logfire tracing
- `index.py`: Webhook endpoint handler
- `tool.py`: Utility functions and webhook processing

Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference