PitLane Agent

AI agent for F1 data analysis using Claude Agent SDK and FastF1.

Quick Start

# Install with pip/uv
pip install pitlane-agent

# Create a workspace
pitlane workspace create --workspace-id my-analysis

# Fetch session data
pitlane fetch session-info --workspace-id my-analysis --year 2024 --gp Monaco --session R

# Analyze lap times
pitlane analyze lap-times --workspace-id my-analysis --year 2024 --gp Monaco --session Q --drivers VER --drivers HAM

Features

• F1 Data Analysis - Lap times, tyre strategy, race telemetry, and more
• Skills-Based Architecture - Specialized skills for different analysis domains
• Workspace Management - Isolated analysis sessions with persistent data
• FastF1 Integration - Direct access to comprehensive F1 data
• Chart Generation - Matplotlib-based visualizations
• Tool Permissions - Domain-restricted web access for safe agent operation

Skills

f1-analyst

Lap time and tyre strategy analysis with visualizations.

Examples:

pitlane analyze lap-times --workspace-id my-analysis --year 2024 --gp Monaco --session Q
pitlane analyze tyre-strategy --workspace-id my-analysis --year 2024 --gp Monaco --session R

f1-drivers

Driver information queries via the Ergast API.

Examples:

pitlane fetch drivers --year 2024
pitlane fetch driver-info --driver-code VER

f1-schedule

Event calendars and session schedules.

Examples:

pitlane fetch schedule --year 2024
pitlane fetch event-info --year 2024 --round 6

Workspace Management

# Create workspace
pitlane workspace create --workspace-id my-analysis

# List workspaces
pitlane workspace list

# Get workspace info
pitlane workspace info --workspace-id my-analysis

# Clean old workspaces (older than 7 days)
pitlane workspace clean --older-than 7

# Remove specific workspace
pitlane workspace remove --workspace-id my-analysis

Requirements

• Python 3.12, 3.13, or 3.14
• Claude API key (for AI agent functionality)

Dependencies

• Claude Agent SDK - AI agent orchestration
• FastF1 - F1 data access and telemetry
• Matplotlib - Chart generation
• Click - CLI framework
• OpenTelemetry - Optional tracing support

Architecture

The agent uses a skills-based architecture where each skill handles a specific F1 analysis domain:

• Skills invoke Python scripts through the agent's tool system
• FastF1 provides data access and telemetry processing
• Matplotlib generates visualizations
• Tool permissions restrict web access to approved F1 data sources

License

Apache-2.0

Links

• GitHub Repository
• Documentation
• Issues