P4 Plan MCP Server

P4 Plan MCP Server is a Model Context Protocol (MCP) server that integrates with P4 Plan. P4 Plan MCP Server exposes safe, structured read/write tools to AI models for P4 Plan.

What you can do with P4 Plan MCP Server

Here are some sample queries you might send to an AI model connected to P4 Plan.

  • "What are my current tasks?"
  • "Show all items tagged to release X"
  • "Show me all bugs with high severity"

Prerequisites

For the most up‑to‑date prerequisites, see the P4 Plan MCP Server README on GitHub.

To use P4 Plan MCP Server, you need:

  • Valid P4 Plan login credentials and a connection to a P4 Plan Server
  • Node.js version 20 or later
  • P4 Plan GraphQL API version 2026.1.002 or later
  • An MCP-compatible client, such as IntelliJ IDEA, Claude Code, or VS Code

Installation

For complete and current installation steps, see the P4 Plan MCP Server README on GitHub.

  1. Download the pre-built binary.

    2. You can also build a P4 Plan MCP Server from source or run it from Docker.

  2. Extract and run the executable.
  3. Configure your MCP client settings and set the access for authenticated users.

    4. LLMs can access anything the authenticated user can. As such, you should be mindful of the authenticated user's access rights.

  4. If necessary, configure P4 Plan to match your MCP client settings.

    5. P4 Plan administrators can manage MCP access through global, group-level, and user-level server properties.

Limitations

Some MCP tools might require non-read-only mode or appropriate P4 Plan permissions. General tool usage is restricted by configurations set by an administrator.

Support and licensing

P4 Plan MCP Server is a community-supported project and is not officially supported by Perforce. Report P4 Plan MCP Server issues to the GitHub community. For P4 Plan issues, see the P4 Plan Support portal.

P4 Plan MCP Server is licensed under the MIT License. To learn more, see the license on GitHub.