Read the full internal state of a WASM agent (messages, turn count, config, stop status). Use when native Task is wrong because the agent runs in a sandboxed WASM runtime whose internal conversation history is not directly accessible from the host process.
AI agents call wasm_agent_state to retrieve information from Claude Flow without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
This is a Read operation: it retrieves data (agent state, conversation history, configuration) without modifying or executing anything. However, severity is elevated to 'medium' because access to internal agent state including full message history and configuration could reveal sensitive information (prompts, user data, API keys in context, internal system directives) that an AI might expose or misuse.
From the tool's definition Tool explicitly states 'Read the full internal state' and retrieves 'messages, turn count, config, stop status' from a sandboxed WASM agent—a query operation with no side effects.
Attacks that exploit this kind of access
Read the full internal state of a WASM agent (messages, turn count, config, stop status). Use when native Task is wrong because the agent runs in a sandboxed WASM runtime whose internal conversation history is not directly accessible from the host process. It is categorised as a Read tool in the Claude Flow MCP Server, which means it retrieves data without modifying state.
Register the Claude Flow MCP server in PolicyLayer and add a rule for wasm_agent_state: 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_state is a Read tool with low risk. Read-only tools are generally safe to allow by default.
Yes. Add a rate_limit block to the wasm_agent_state 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_state. 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_state 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.