The Agent Skills Directory

[SAFE]: The skill implements a set of specialized tools (mcp__plugin_ruflo-core_ruflo__wasm_agent_*) dedicated to creating, prompting, and managing isolated WASM environments. This architecture is inherently security-focused as it targets code sandboxing.
[COMMAND_EXECUTION]: The Bash tool is listed in the allowed-tools frontmatter, which is a platform-level restriction. The skill instructions do not invoke Bash for any sensitive or dangerous operations; its inclusion provides the agent with capability without presenting an active threat in the provided context.

wasm-agent