seo-sitemap

Pass

Audited by Gen Agent Trust Hub on Apr 17, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill processes untrusted XML sitemap data, creating a surface for indirect prompt injection attacks where data could influence agent behavior.
  • Ingestion points: Data enters the context in 'Mode 1' when the skill analyzes existing sitemaps from external files or URLs.
  • Boundary markers: The instructions do not specify the use of delimiters or 'ignore' commands to isolate the ingested data from the agent's core instructions.
  • Capability inventory: The skill utilizes tool capabilities for network requests to validate URL status codes and file system access to generate reports and new sitemap files.
  • Sanitization: There are no explicit requirements for sanitizing the ingested sitemap content to prevent the execution of embedded instructions.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 17, 2026, 02:09 PM
Security Audit — agent-trust-hub — seo-sitemap