ads-dna
Pass
Audited by Gen Agent Trust Hub on Jun 16, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: Fetches HTML and CSS content from arbitrary user-provided URLs using
agent_browser,curl, orPlaywright. This is an essential function for brand analysis but involves retrieving data from untrusted sources. - [COMMAND_EXECUTION]: Provides instructions for installing browser dependencies using
python3 -m playwright install chromiumif automated capture fails. - [PROMPT_INJECTION]: Processes untrusted data from external websites, creating a surface for indirect prompt injection. Malicious content on a scanned site could attempt to influence agent output or the resulting brand profile.
- Ingestion points: Web page content retrieved from user-specified URLs.
- Boundary markers: No specific delimiters or "ignore" instructions are defined for the fetched content during analysis.
- Capability inventory: Local file system writes (
brand-profile.json) and browser automation. - Sanitization: Fetched text is analyzed without explicit sanitization or filtering of potential injection patterns.
Audit Metadata