agent-wasm-specialist
Pass
Audited by Gen Agent Trust Hub on Oct 2, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONNO_CODE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill interpolates external data into its final task instructions using the
$ARGUMENTSvariable, which creates a vulnerability surface for indirect prompt injection attacks. - Ingestion points: External data is ingested through the
$ARGUMENTSplaceholder at the end of theSKILL.mdfile. - Boundary markers: The skill lacks protective delimiters (such as XML tags or unique markers) or specific system instructions to treat the interpolated text strictly as data rather than instructions.
- Capability inventory: The agent is configured as a specialist for WebAssembly compilation and runtime optimization, roles that typically require command-line access, file system manipulation, and toolchain execution.
- Sanitization: No input validation, escaping, or filtering is applied to the content before it is processed by the agent.
- [NO_CODE]: This skill consists exclusively of informational markdown and configuration metadata. It does not ship with any executable scripts, binaries, or configuration files that install external dependencies.
Audit Metadata