Skip to main content
Glama
iRacing MCP Server

iRacing MCP Server

by ellettie

iRacing MCP Server

An MCP (Model Context Protocol) server for accessing iRacing telemetry data and game functionality.

Overview

This project provides real-time telemetry data, leaderboard information, camera control, pit commands, replay functionality, and more from iRacing through the MCP protocol. It enables AI assistants and other applications to easily access iRacing data.

Related MCP server: @racehooks/mcp

Key Features

📊 Telemetry Data

  • Retrieve real-time telemetry values

  • Get list of available telemetry variables

🏁 Race Information

  • Get leaderboard (competitive positions only)

  • Access driver information, session information, weekend information, and split time information

  • Monitor current flags and engine warnings

📹 Camera Control

  • Get available camera groups

  • Switch cameras (by car number, position, group specification)

  • Check current camera status

🔧 Pit Operations

  • Check pit service status

  • Execute pit commands (refueling, tire changes, repairs, etc.)

  • Manage safe pit operations

🎬 Replay Features

  • Search and navigate replays

  • Move between sessions, laps, and frames

  • Jump to incident markers

Requirements

  • Python 3.13 or higher

  • iRacing (must be running)

  • uv

Usage

1. Configure mcp.json

{
    "mcpServers": {
        "iracing-mcp-server": {
            "command": "uvx",
            "args":["iracing-mcp-server"]
        }
    }
}

Available Tools

  • get_telemetry_names() - Get available telemetry variables

  • get_telemetry_values(names) - Get specified telemetry values

Race Information

  • get_leaderboard() - Get leaderboard

  • get_driver_info() - Get driver information

  • get_session_info() - Get session information

  • get_qualify_results_info() - Get qualification results information

  • get_weekend_info() - Get weekend information

  • get_split_time_info() - Get split time information

  • get_radio_info() - Get radio infomation

  • get_current_flags() - Get current flags

  • get_current_engine_warnings() - Get engine warnings

Camera Control

  • get_camera_info() - Get camera information

  • get_current_camera_status() - Get current camera status

  • cam_switch(group_number, car_number_raw, position) - Switch camera

Pit Operations

  • get_current_pit_service_status() - Get current pit service status

  • pit_command(commands_and_values) - Execute pit command

Replay Features

  • replay_search(search_commands) - Search and navigate replay

Development

Dependencies

  • mcp[cli]>=1.12.2 - MCP protocol implementation

  • pyirsdk>=1.3.5 - iRacing SDK Python bindings

Project Structure

iracing-mcp-server/
├── src/iracing_mcp_server/
│   ├── __init__.py          # Main entry point
│   ├── server.py            # MCP server implementation
│   ├── leaderboard.py       # Leaderboard processing
│   └── prompt.py            # Prompt templates
├── pyproject.toml           # Project configuration
└── README.md               # This file

License

This project is released under the MIT License.

Acknowledgments

  • pyirsdk - Python bindings for iRacing SDK

  • MCP - Model Context Protocol

Available Tools

17 tools
cam_switchD

switch camera

Inputs: group_number, car_number_raw, position. Returns: result

get_camera_infoC

get camera info

Returns: list[dict]: camera info

Inputs: no named parameters. Returns: result

get_current_camera_statusB

get current camera status

Returns: dict: current camera status - target_car_idx: int - camera_group_number: int - camera_number: int

Inputs: no named parameters.

get_current_engine_warningsA

get current engine warnings

Returns: list[str]: current engine warnings

Inputs: no named parameters. Returns: result

get_current_flagsD

get current flags

Returns: list[str]: current flags

Inputs: no named parameters. Returns: result

get_current_pit_service_statusC

get current pit service status

Returns: list[str]: current pit service status

Inputs: no named parameters. Returns: result

get_driver_infoC

get driver info

Returns: dict: driver info

Inputs: no named parameters.

get_leaderboardB

get leaderboard

Note: When filtering for competitive leaderboard results, exclude entries where: 1. if driver_name="Pace Car" (not a competitor. so ignore this row from leaderboard) 2. if is_missing_start=True (indicates missed start. so this car is not correct position) 3. if is_towing=True (indicates towing. so this car is not correct position) 4. if status="not_in_world" (not in world. so this car is not correct position)

Returns: list: leaderboard each element is a dict with the following keys: - car_idx: int - position: int | None - class_position: int | None - car_number: int - car_number_raw: int (use this value for camera switching functions) - driver_name: str - team_name: str - irating: int - license_str: str - lap_and_lap_dist_pct: float - best_lap_time: float - last_lap_time: float - car_name: str - class_name: str - class_est_time: float - gap_to_leader_str: str - gap_to_next_str: str - gap_to_class_leader_str: str - gap_to_class_next_str: str - flags: list[str] (e.g. ["black", "repair", "furled", "servicible"]. "servicible" means no flags) - status: str (e.g. "off_track", "in_pit_stall", "approaching_pits", "on_track", "not_in_world") - is_missing_start: bool - is_towing: bool

Inputs: no named parameters.

get_qualify_results_infoD

get qualify results info

Returns: list[dict]: qualify results info

Inputs: no named parameters. Returns: result

get_radio_infoD

get radio info

Returns: dict: radio info

Inputs: no named parameters.

get_session_infoD

get session info

Returns: dict: session info

Inputs: no named parameters.

get_split_time_infoB

Retrieves split time information for each sector of the track.

Returns: dict: split time info

Inputs: no named parameters.

get_telemetry_namesD

get telemetry names

Returns: list[str]: telemetry names

Inputs: no named parameters. Returns: result

get_telemetry_valuesA

get telemetry values. to get available telemetry names, call get_telemetry_names. if you want to get all telemetry values, set names to None.

Returns: dict: telemetry values

Inputs: names.

get_weekend_infoD

get weekend info

Returns: dict: weekend info

Inputs: no named parameters.

pit_commandC

apply pit command. before apply pit command, you should call get_current_pit_service_status.

Inputs: commands_and_values (required). Returns: result

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

    TDQS

    C2.5/5.0

    Scored across 17 tools

    Disambiguation4/5

    Most tools have clearly distinct purposes, but the camera-related tools (cam_switch, get_camera_info, get_current_camera_status) could cause confusion for an agent due to overlapping focus.

    Naming Consistency3/5

    The majority of tools follow a 'get_*' pattern, but there are exceptions like 'cam_switch', 'pit_command', and 'replay_search' that break the convention, making it somewhat inconsistent.

    Tool Count5/5

    With 17 tools, the set is well-scoped for an iRacing simulation server. Each tool serves a clear purpose without being excessive or insufficient.

    Completeness4/5

    The tool surface covers a broad range of simulation aspects (session, driver, leaderboard, telemetry, etc.), with only minor gaps such as missing dedicated live timing updates or additional race control commands.

    Maintenance

    ActivityInactive
    ResponsivenessNo recent community issues

    Related MCP Connectors

    Related MCP Servers

    • A
      license
      B
      quality
      B
      maintenance
      MCP server that gives AI assistants direct access to RaceHooks motorsport analytics and data, enabling management of webhooks, live sessions, simulations, and account usage.
      58
      101 npm
      1
      MIT
    • F
      license
      Not graded
      quality
      D
      maintenance
      MCP server that provides Forza Horizon 5 telemetry tools (session listing, summaries, telemetry windows, tuning recommendations) to AI assistants, enabling natural language queries about driving sessions.
      -
    • A
      license
      A
      quality
      B
      maintenance
      Gives AI assistants structured access to racing-game telemetry such as lap and sector times, lap comparisons, braking zones and tyre wear through MCP tools and resources. It works out of the box on bundled synthetic sample data and can ingest a live F1 23 UDP feed.
      7
      MIT