Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseDotsDots+21 more

    gitbash mcp

    1

    Execute Git Bash commands, manage background jobs, and handle human-in-the-loop approvals on Windows.

    Free

    Security scanned
    gitbash mcp

    gitbash mcp

    Example session with this skill installed

    Run a git status in C:\projects\web-app and if it's clean, start a production build in the background.

    • Read your context and instructions
    • Compiled the gitbash mcp

    I have checked the repository status. The working directory is clean, so I have initiated the production build as a background job.

    Background Execution Started

    FieldValue
    Commandgit status --short && npm run build
    DirectoryC:\projects\web-app
    Job IDjob_8f2a1b9c3e
    Statusrunning
    Policy Modeask

    Next steps

    • Run job_output {"job_id": "job_8f2a1b9c3e", "wait": true} to check the build progress.
    • Use job_list {} to see all active background processes.
    • If the build hangs, use job_kill {"job_id": "job_8f2a1b9c3e"}.

    Connects securely to your tools. The creator never sees your data.

    About this skill

    Windows developers often struggle to bridge the gap between AI agents and local development environments. This skill enables seamless execution of Git Bash commands, CI/CD scripts, and complex Git workflows directly on Windows machines via MCP.

    What it does

    • Execute Bash scripts runs multi-line scripts or one-off commands directly in a Git Bash environment.
    • Manage background jobs handles long-running builds or tests with persistent job tracking and output polling.
    • Control command policy manages safety levels via ask or allow modes for risky operations.
    • Manage human approvals coordinates with a local Windows WPF interface for manual verification of sensitive actions.
    • Diagnose environment issues uses built-in tools to verify Node.js, Git Bash, and configuration health.

    How it works

    1. Initialize the server by running gitbash-mcp init targeted at your specific IDE or client like Cursor or Claude Desktop.
    2. Execute commands using the exec tool, providing the bash command and a specific Windows cwd path.
    3. Track background work by requesting a job_id for long processes and polling for stdout/stderr increments.
    4. Manage approvals via the local Windows window for commands flagged by the security policy.

    Frameworks & tools

    Built for Windows environments requiring Node.js 18+. Integrates Git Bash, Windows PowerShell, and WPF for human-in-the-loop interfaces. Supports DSH, Claude Code, Cursor, and VS Code.

    Why this beats prompting it yourself

    Raw LLMs struggle with Windows path conversion and process management. This skill provides a structured interface for job recovery, human-in-the-loop safety, and persistent background execution that standard terminal prompts cannot maintain.

    Use cases

    • Automate Git workflows like complex rebases, branch cleanup, or multi-repo status checks.
    • Run local builds and test suites as background jobs without blocking the agent's chat interface.
    • Set up dev environments by installing dependencies and configuring local environment variables.

    Known limitations

    Requires Windows, Node.js 18+, and Git Bash. No filesystem sandbox is provided. Global installation via npm is required. Environment variables do not persist across separate exec calls.

    Requires Windows, Node.js 18+, and Git Bash. No filesystem sandbox provided. Shell state like variables and cd do not persist between separate tool calls.

    How to use

    1. 1

      Connect your agent to the Agensi MCP

      Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup

    2. 2

      Sign in to Agensi

      Your agent shows a sign-in link. Sign in and click Allow.

    3. 3

      Ask for what you need

      Your agent loads this skill and uses it on your task. It always asks before any payment.

    Reviews

    No reviews yet

    Be one of the first to try it. Every listed skill goes through the trust checks below.

    Security scanned

    Checked by our 8-point scan before listing

    Fresh listing

    Recently published to Agensi

    Free forever

    No account required to browse

    Trust & safety

    Security scan

    Security scan passed 10 days ago

    • Free with an Agensi account

    Requires

    git
    Node.js
    or

    Permissions

    Terminal / Shell
    Read Files
    Environment Variables
    Write Files
    Network Access
    Browser
    Listed10 days ago

    What's inside

    Frequently Asked Questions

    This skill bridges the gap between AI agents and Windows environments by allowing the agent to execute complex Git Bash scripts, handle background build processes, and interact with a local WPF interface for manual command approvals.

    It is compatible with any MCP-enabled client, including Claude Desktop, Cursor, VS Code, and Claude Code, provided the host machine is running Windows with Node.js 18+ and Git Bash installed.

    Connect your AI agent to Agensi through MCP, then ask it to use gitbash mcp on your task. Your agent loads the latest approved version each time. Paid skills require a purchase or active subscription. Skill files are not available for download; your licence does not allow you to save, copy or export any part of a skill.

    This skill includes a robust security policy system that allows you to toggle between 'ask' and 'allow' modes, alongside a dedicated Windows popup window for manual human-in-the-loop verification of sensitive commands.

    Yes, unlike standard terminal prompts, this skill provides persistent job tracking that allows you to start a long-running process, receive a job ID, and poll for output updates without blocking the agent's other functions.

    New in LLM Ops