$ loading_
Run commands, manage long jobs, and transfer files in AI sandboxes.
Copy the install command and let the AI configure it · recommended for beginners
No copy-paste install info for "sandbox-mcp" yet — see the docs or source repo.
Run the following commands in an isolated sandbox: git status && pytest -q. Return the full output, an error summary, and suggested next fixes.
Command results, a summary of test failures, and actionable fix recommendations.
Start a background training job in the Docker sandbox: python train.py --epochs 50. Record the job ID, provide periodic log summaries, and return the output file list when finished.
A job ID, periodic status updates, log summaries, and a final list of generated artifacts.
Transfer a local data archive into the sandbox, extract it, count rows and schema for each CSV file, and return the results in a table.
File transfer and processing status, plus a structured table of row counts and column schemas for each CSV.
Safely run any code in isolated Docker containers for testing and automation.
Give AI agents authenticated sandbox access and a restricted path for safe production deploys.
A sandbox server for testing and debugging MCP tools and interactions.
Securely equips AI agents with executable tools for commands, search, and file operations.
Safely read, write, search, and manage files in a sandboxed environment.
Verify MCP server responses by returning a unique canary identity string.