Plimsoll

What is Plimsoll

Plimsoll is a self-hosted code sandbox designed for securely running code generated by AI Agents.

Key features

• Safely executes untrusted code generated by AI Agents inside the sandbox
• Each run result is clearly labelled with an isolation strength level (process, container, kernel, vm)
• Requests can specify a minimum acceptable isolation strength; execution is refused if the requirement is not met
• Built-in dry-run mechanism that subjects AI-generated code to scenario-based testing and returns a SHA-256 hash for verification
• Provides the plimsoll-attest tool to sign execution results, verify chain integrity and replay them

Who it is for

For developers, researchers and teams who need to securely execute code generated by AI Agents, suitable for scenarios with explicit requirements on code isolation strength.

How to get started

1. Visit https://github.com/plimsollmark/plimsoll to read the project documentation
2. Compile and run the plimsoll daemon following the instructions in the README
3. Use the official Go client to make requests with specified isolation strength requirements
4. Sign and verify run results using the plimsoll-attest tool

Compiled from public descriptions on the tool's official website / README / launch page. Pricing and user numbers have not been verified; please check the official website. Some sections were drafted with AI assistance. Written in Chinese and English; other languages are machine-translated.