platform-voice-check
Pass
Audited by Gen Agent Trust Hub on Sep 3, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: No malicious patterns, such as credentials exfiltration or unauthorized system access, were detected. The skill operates as a controlled text-processing gate within a specified workflow.
- [INDIRECT_PROMPT_INJECTION]: The skill processes external text files, which constitutes an ingestion surface for untrusted data.
- Ingestion points: The workflow reads four output files (e.g.,
substack-note.md) and external voice profiles for auditing. - Boundary markers: The instructions do not define delimiters or specific markers to isolate audited content from instructions.
- Capability inventory: The skill is limited to reading files and writing markdown reports; it lacks access to shell execution, network tools, or dynamic code evaluation.
- Sanitization: No explicit logic is provided to sanitize or ignore embedded instructions within the audited documents.
Audit Metadata