synthesis-voice-profiler
Pass
Audited by Gen Agent Trust Hub on Oct 11, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted user writing samples from various sources (text, local files, or external URLs) to generate instructions for high-privilege configuration files like CLAUDE.md.
- Ingestion points: Writing samples provided by the user in Step 1.
- Boundary markers: The instructions do not define delimiters or provide warnings to the agent to ignore embedded instructions within the samples.
- Capability inventory: The skill uses the agent's fetching and file-reading capabilities to collect data for analysis and generates markdown profiles. It does not invoke shell commands or transmit data to external servers.
- Sanitization: There is no explicit sanitization of the processed samples, although the process includes a human review step in Step 5.
Audit Metadata