cast
Pass
Audited by Gen Agent Trust Hub on Oct 22, 2026
Risk Level: SAFEREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [REMOTE_CODE_EXECUTION]: The skill uses
npx --yes edge-ttsto list and verify available voices. This operation downloads and executes the package from the npm registry at runtime. - [COMMAND_EXECUTION]: The
SPEAKengine performs environment checks using shell commands such aswhichto locate binaries andpython3to verify the presence of the Google Cloud Text-to-Speech library. - [INDIRECT_PROMPT_INJECTION]: The
CONJUREandFUSEworkflows process data from repository files includingREADME.md, source code, and user feedback. Because these sources are untrusted, they could contain embedded instructions intended to influence the persona synthesis process. The skill attempts to mitigate this with confidence scoring and evidence-grounding requirements. - [EXTERNAL_DOWNLOADS]: The skill retrieves and references guidelines and research data from well-known sources, including the npm registry, Nielsen Norman Group, and academic repositories like arXiv and the ACM Digital Library.
Audit Metadata