runeward MCP
v0.1.4io.github.runewardd/runeward
Governed execution cells for AI agents: Docker/K8s sandboxes with policy, audit, and approvals.
transport:stdio
runtime:oci
Target client
run in your project directory
claude mcp add runeward -e RUNEWARD_CONFIG_DIR=<RUNEWARD_CONFIG_DIR> -e RUNEWARD_STATE_DIR=<RUNEWARD_STATE_DIR> -- docker run -i --rm ghcr.io/runewardd/runeward:v0.1.4
Adds it for this project only. Append --scope user to make it available everywhere. Claude Code docs
This listing does not declare its tools. Connect the server and your client will discover them on the handshake.