weak-internal-links
Pass
Audited by Gen Agent Trust Hub on Oct 7, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill requires the agent to crawl external websites and review rendered HTML content, which introduces a surface for indirect prompt injection attacks where malicious instructions could be embedded in the audited site's content.
- Ingestion points: The skill instructions in
SKILL.md("Crawl the site and build an internal link graph") andreferences/rule.md("Inspect rendered HTML and HTTP headers") direct the agent to process data from external, untrusted sources. - Boundary markers: No explicit boundary markers or delimiters are defined to separate the instructions from the external content being analyzed.
- Capability inventory: The skill utilizes network access for crawling and high-level HTML/metadata analysis capabilities.
- Sanitization: There are no instructions for sanitizing or escaping the data retrieved from external websites before the agent processes it.
Audit Metadata