sentiment-analyzer
Pass
Audited by Gen Agent Trust Hub on Nov 9, 2026
Risk Level: SAFE
Full Analysis
- [DATA_EXFILTRATION]: The skill processes media files and transcripts locally or via official API calls. Sensitive data like API keys are handled via environment variables, which is a secure practice. No evidence of unauthorized data harvesting or exfiltration to untrusted domains was found.
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted transcript data as an input surface. However, this data is primarily used for keyword matching and as content for sentiment analysis, and it is not used to construct executable commands or bypass safety constraints.
- [DYNAMIC_EXECUTION]: The script uses runtime imports to load local utility modules from the shared project environment. These are standard architectural components and do not involve the dynamic execution of untrusted external code or remote scripts.
Audit Metadata