Execute a tool on a WASM agent sandbox. Tools: read_file, write_file, edit_file, write_todos, list_files. Use flat format: {tool, path, content, ...}. Use when native Task is wrong because the workload needs sandboxed isolation — untrusted code execution, browser-side run, deterministic replay. P...
AI agents invoke wasm_agent_tool to trigger actions in Claude Flow. What it does depends on the arguments the agent supplies, and its effects often reach beyond the immediate call — builds kicked off, notifications sent, workflows started.
This tool executes arbitrary operations (read_file, write_file, edit_file, list_files) within a WASM sandbox environment. The description explicitly mentions 'untrusted code execution' and 'browser-side run' as use cases, indicating it triggers external operations whose effects depend on arguments.
From the tool's definition Execute a tool on a WASM agent sandbox. Tools: read_file, write_file, edit_file, write_todos, list_files. Use when native Task is wrong because the workload needs sandboxed isolation — untrusted code execution, browser-side run, deterministic replay.
Attacks that exploit this kind of access
Execute a tool on a WASM agent sandbox. Tools: read_file, write_file, edit_file, write_todos, list_files. Use flat format: {tool, path, content, ...}. Use when native Task is wrong because the workload needs sandboxed isolation — untrusted code execution, browser-side run, deterministic replay. Pair with wasm_gallery_search to find a published agent, or wasm_agent_create to scaffold a fresh one. For trusted in-process work, native Task is fine. It is categorised as a Execute tool in the Claude Flow MCP Server, which means it can trigger actions or run processes. Use rate limits and argument validation.
Register the Claude Flow MCP server in PolicyLayer and add a rule for wasm_agent_tool: allow, deny, rate-limit, or require approval. Point your MCP client at the PolicyLayer proxy URL and the rule is enforced on every call, before it reaches Claude Flow. Nothing to install.
wasm_agent_tool is a Execute tool with high risk. Execute tools should be rate-limited and have argument validation enabled.
Yes. Add a rate_limit block to the wasm_agent_tool rule in your PolicyLayer policy. For example, setting max: 10 and window: 60 limits the tool to 10 calls per minute. Rate limits are tracked per agent session and reset automatically.
Set action: deny in the PolicyLayer policy for wasm_agent_tool. The AI agent will receive a policy violation error and cannot call the tool. You can also include a reason field to explain why the tool is blocked.
wasm_agent_tool is provided by the Claude Flow MCP server (claude-flow). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.