speech-build

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS] (LOW): The skill references source code on GitHub for the official Google GenAI SDK. Evidence: references/source_code.md contains links to raw.githubusercontent.com/googleapis/python-genai/. Per the [TRUST-SCOPE-RULE], this is a trusted source.
  • [PROMPT_INJECTION] (LOW): The skill utilizes a natural language interface to control speech output (e.g., Director's Notes), creating a surface for indirect prompt injection. 1. Ingestion points: 'contents' parameter in generate_content (SKILL.md). 2. Boundary markers: Markdown headers like '#### TRANSCRIPT' are suggested as delimiters. 3. Capability inventory: Generation of audio from text. 4. Sanitization: No explicit sanitization of input text is documented.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 06:07 PM