- READY
- VERIFIED
- LOCAL
- HUMAN
Short answer
How can my AI pull and prep code for me?
Your AI uses Dagger to clone remote Git repos and initialize them inside a container. This gives you a ready-to-run environment without manual setup. You'll see the exact steps for fetching, inspecting, and containerizing your code below.
The results of a code pull.
Where your code lands.
What happens after you give the command.
READY
Initialized containers
The AI pulls the repo and sets up the directory structure. You get a container ready for execution.
VERIFIED
Directory inspection
The agent checks the files and folder structure to ensure everything is where it should be before running.
LOCAL
Local file access
The AI maps the pulled code to a specific directory, making it visible for subsequent tasks.
HUMAN
Manual review needed
You decide if the pulled code or the container environment meets your specific requirements before execution.
The workflow
What your AI does when the code arrives.
The agent moves from a remote URL to a running container environment.
Fetch the repository
The agent reaches out to the remote Git provider to clone the specific repository you requested.
query_gitInspect the files
Once cloned, the agent looks through the file tree to understand the project structure.
query_directoryPrepare the environment
The agent maps the files into a containerized space to ensure a clean execution environment.
query_containerValidate the setup
The agent checks the container state to confirm the directory is ready for your commands.
query_containerquery_directory
Try it
Copy these to start.
Direct your agent to begin the setup process.
Starting points
Swap the repository URLs and directory paths with your actual project details.
Dagger Connector
You're all set. Choose your MCP client and follow the setup instructions.
https://edge.vinkius.com/vk_preview_jU749naeDKubVrRR38BTiJPB8XzxtT6YXppJtFAW/mcpClaude Desktop
Follow the steps below to connect in seconds.
- 1In Claude Desktop, open Settings → Connectors.
- 2Click “Add custom connector” and paste the connector link above as the remote MCP server URL.
- 3Click Add and start a new chat — Dagger capabilities are ready to use.
{
"mcpServers": {
"dagger-programmable-ci-mcp": {
"url": "https://edge.vinkius.com/vk_preview_jU749naeDKubVrRR38BTiJPB8XzxtT6YXppJtFAW/mcp"
}
}
}Pull the repository at https://github.com/example/repo and prepare it for execution.
Clone the code from the remote Git repo and show me the directory structure.
Initialize a container with the files from the project in /workspace/my-app.
Fetch the source code from the Git URL and verify the files are in the container.
Claude
ChatGPT
Cursor
VS Code
Windsurf
Claude Code
JetBrains
Cline
Start here
Connect Dagger once, then ask.
One connection gives your agent access to these tools. Your credentials stay encrypted and secure on our side.
Connect Dagger to your AIFAQ
How this task behaves.
- 01
Can the AI push code back to GitHub?
No. This MCP only has tools to query Git repositories, not to write or push changes to them.
- 02
Does the AI delete my local files?
No. The agent works with remote repositories and containerized directories, leaving your local machine untouched.
- 03
Can I run arbitrary shell commands?
No. The agent can only perform the specific actions defined by the query tools in the manifest.
- 04
How does the AI know if the code is in the container?
The agent uses the container tool to inspect the environment and confirm the files exist where expected.
- 05
Can the AI change my container configuration?
No. The agent can query and inspect the container, but it cannot modify the underlying container image or settings.
More questions about Dagger? The Connector page answers them. See everything the Dagger Connector can do
Connect Dagger to Claude, Cursor, ChatGPT & more
