seo-content-brief

Pass

Audited by Gen Agent Trust Hub on Oct 4, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from external sources (Google SERP, competitor websites, and general search results) using tools such as dataforseo_v3_serp_google_organic_live_advanced, context_dev_scrape_markdown, exa_search, and tavily_search. This creates an attack surface where malicious instructions embedded in web content could attempt to influence the agent's behavior at runtime.
  • Ingestion points: Data is retrieved via search and scraping tools listed in SKILL.md and references/sandbase-api-map.md.
  • Capability inventory: The agent can search the web and scrape markdown content, but does not have file-write or shell execution capabilities documented in this skill.
  • Boundary markers: There are no explicit instructions or delimiters defined to separate untrusted web content from agent instructions, which increases the likelihood of the agent following instructions found within the data.
  • Sanitization: The instructions do not specify any sanitization or validation of the content retrieved from external searches.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 4, 2026, 11:45 AM