seo-internal-linking
Pass
Audited by Gen Agent Trust Hub on Sep 29, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill describes a workflow that involves crawling external websites ('Crawl site (cheerio + breadth-first)') to perform audits. This ingestion of untrusted data from the web represents a potential surface for indirect prompt injection if the audited content contains malicious instructions designed to influence the agent's behavior during the audit process.
- Ingestion points: Workflow step 1 ('Crawl site') involves fetching external HTML content.
- Boundary markers: None specified in the instructions to delimit external content from system instructions.
- Capability inventory: The workflow implies network access (crawling) and data processing (graph building, orphan detection).
- Sanitization: No explicit sanitization or filtering of crawled content is mentioned.
Audit Metadata