Plimsoll is a self-hosted sandbox service for running untrusted, agent-authored code. Its distinguishing detail is that each run reports back the isolation boundary it actually executed behind, so you can see what protected you instead of assuming it. It also ships browser-based lessons that need no clone to try.
Be the first to rate Plimsoll. Your feedback helps other users decide.
Execute code an agent wrote without trusting it on your host
See which isolation layer each run actually landed behind
Work through lessons without cloning or installing anything
Similar tools in the same category — compare side-by-side
Corral is an open-source utility for killing every command your AI agent starts. When an autonomous ...
DwarfStar 4 (ds4) is a narrow C inference engine for running frontier open-weight models locally on ...
MemorySync gives AI agents persistent memory through an API. In the browser playground you tell it o...
Guides and comparisons about code tools
Ten October 2026 agent infrastructure launches: local inference engines, sandboxes that report their isolation boundary, agent memory and MCP testing.
Nine tools that constrain what an agent can produce and reach: slop-detecting lint rules, declared sandbox tiers, cheaper crawling and signed messaging.
Nine Show HN launches beneath the model: filesystems, snapshots, gateways and deploy tools that decide what an agent sees and what survives.
Nine launches that check agent work: per-call access policy, MCP server vetting, runtime verification, readable diffs and a hard stop at 95%.
Nine Show HN launches sitting under the agent: a model gateway, vectorless retrieval, on-device inference, an MCP browser layer and connectors for real systems.
