voice-refine
Pass
Audited by Gen Agent Trust Hub on Oct 12, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill's primary function is to transform untrusted, user-provided voice transcriptions into structured prompts, which creates a surface for indirect prompt injection.
- Ingestion points: The skill accepts unstructured text inputs (transcriptions) from the user for the refinement pipeline defined in SKILL.md.
- Boundary markers: The instructions do not specify the use of clear delimiters or instructions to ignore potential commands within the user input, which could lead the agent to follow malicious instructions hidden in a transcript.
- Capability inventory: The skill is restricted to the 'Read' tool in its YAML frontmatter, which significantly limits the potential for harm if an injection were to occur.
- Sanitization: While the skill defines rules to filter filler words and hedging, it does not include sanitization or validation logic to detect or strip executable commands or malicious prompt overrides from the input data.
Audit Metadata