wasm-agent
Pass
Audited by Gen Agent Trust Hub on May 19, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [SAFE]: No malicious patterns were detected. The skill implements a security-focused architecture by using WebAssembly (WASM) to isolate agent execution from the host system.
- [PROMPT_INJECTION]: The skill includes an attack surface for indirect prompt injection because it allows external data to be passed directly into an agent's execution context.
- Ingestion points: The
mcp__claude-flow__wasm_agent_prompttool inSKILL.mdis used to send untrusted prompts to the WASM agent. - Boundary markers: The instructions do not define delimiters or specific "ignore embedded instructions" headers for the data passed to the agent.
- Capability inventory: The WASM environment has capabilities for file manipulation (
mcp__claude-flow__wasm_agent_files) and tool execution (mcp__claude-flow__wasm_agent_tool) within its isolated sandbox. - Sanitization: No input sanitization or validation steps are mentioned for the content provided to the WASM agent.
Audit Metadata