meta-ads-library
Pass
Audited by Gen Agent Trust Hub on May 5, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection.\n * Ingestion points: Ad copy is scraped from the Meta Ads Library via
scripts/scrape.py.\n * Boundary markers: The generatedsummary.mdandresearch-brief.mdfiles include scraped ad text without clear delimiters or instructions to ignore embedded commands. This could allow malicious instructions in ad copy to influence the agent's behavior when it later processes these reports.\n * Capability inventory: The skill utilizes scripts for browser automation (scripts/scrape.py) and file system operations (scripts/normalize.py,scripts/aggregate.py).\n * Sanitization: No sanitization or escaping of the scraped ad copy is performed before it is written to markdown reports.\n- [COMMAND_EXECUTION]: The skill provides Python scripts that execute shell commands and automate browser actions using the Playwright library to scrape and process data.
Audit Metadata