udemy-crawler
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill creates an attack surface by ingesting untrusted data from external Udemy course pages, which could theoretically contain instructions intended to influence the agent's formatting behavior.\n
- Ingestion points: Scrapes
udemy.comcourse headlines, descriptions, and curriculum content inSKILL.md.\n - Boundary markers: The skill does not define specific delimiters or warnings to separate scraped content from agent instructions in its output template.\n
- Capability inventory: Capabilities are limited to browser automation and text formatting; there is no evidence of file-system write access or network exfiltration based on the scraped content.\n
- Sanitization: Scraped text is interpolated directly into a markdown template without specific escaping or filtering.\n- [COMMAND_EXECUTION]: The skill relies on the execution of shell commands to run browser automation scripts.\n
- Evidence: Provides
node evaluate.jscommand blocks inSKILL.mdto initiate scraping tasks.
Audit Metadata