Wasm Agent

This lets you spin up Claude agents that run in WebAssembly sandboxes, which means they can't touch your actual filesystem or do anything destructive to your system. You create an agent with a config, send it prompts, give it specific tools, and manage files within its isolated environment. The main win here is safety when you're experimenting with agent behaviors or running code you don't fully trust. You can also export the whole thing as a portable package that runs anywhere WASM does. It's overkill if you're just doing basic scripting, but if you're building agent systems where isolation actually matters or you need reproducible execution across different environments, it solves that problem cleanly.

skills.sh installs: 740
GitHub stars: 81.4K
last updated: Aug 21, 2026

View on GitHub

Install

$ npx -y skills add ruvnet/ruflo --skill wasm-agent --agent claude-code

Installs into .claude/skills of the current project.

Files

WASM Agent

Create sandboxed agents that run in WebAssembly for safe, isolated execution.

When to use

When you need to run untrusted code, experiment with agent configurations, or create portable agents that run anywhere WASM is supported.

Steps

Create agent — call mcp__plugin_ruflo-core_ruflo__wasm_agent_create with agent configuration
Send prompt — call mcp__plugin_ruflo-core_ruflo__wasm_agent_prompt to interact with the agent
Use tools — call mcp__plugin_ruflo-core_ruflo__wasm_agent_tool to give the agent access to specific tools
Manage files — call mcp__plugin_ruflo-core_ruflo__wasm_agent_files to read/write files in the sandbox
Export — call mcp__plugin_ruflo-core_ruflo__wasm_agent_export to package the agent for sharing
List agents — call mcp__plugin_ruflo-core_ruflo__wasm_agent_list to see all running WASM agents
Terminate — call mcp__plugin_ruflo-core_ruflo__wasm_agent_terminate to stop an agent