add-dictation
Pass
Audited by Gen Agent Trust Hub on Sep 19, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill facilitates the ingestion of untrusted audio data, which is converted to text and then inserted into the application's input field.
- Ingestion points: Transcribed text is received from the Grok API endpoints
https://api.x.ai/v1/sttandwss://api.x.ai/v1/sttas documented inSKILL.md. - Boundary markers: The provided implementation examples do not use delimiters or instructions to treat the resulting transcript as untrusted data.
- Capability inventory: The
insertAtCursor(text)function (mentioned inSKILL.md) places the potentially malicious transcription into the user's active composer. - Sanitization: There is no evidence of sanitization or filtering of the transcribed text before it is inserted into the UI context.
Audit Metadata