⬢github Go · 25 ★ +23 since we first saw it · pushed 7 h ago · Apache-2.0
plimsollmark/plimsoll
A sandbox service for running untrusted, agent-authored code that reports the isolation boundary each run executed behind. Lessons, no clone needed: https://plimsollmark.github.io/plimsoll/trainers/
Plimsoll is a self-hosted Go service that runs untrusted, agent-authored code in a sandbox and reports which isolation tier (process, container, kernel, or vm) actually protected each run. Requests can set a minimum tier, results carry SHA-256 hashes for tamper evidence, and a companion tool signs and replays run records.
Why now: It was just shared on Hacker News as a 'Show HN' demo, with a live cart-pole run report showing how the sandbox grades agent-written code.
Who it is for: Developers building LLM agents or platforms who need to execute AI-generated code safely and verifiably.
ai-agentscode-executioncode-interpretere2bfirecrackergolanggvisorllm-agentsmcpsandbox
Stars over our 126 snapshots: 2 to 25, since 1 d ago.
Where people talked about it
API: https://socialmediatrends-api.osmike.com/v1/repos/plimsollmark/plimsoll