robots-ai-crawler-audit

Pass

Audited by Gen Agent Trust Hub on Sep 3, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill provides a comprehensive set of diagnostic shell command templates using curl, awk, sed, and dig to perform network-based audits of a target website's robots.txt, security headers, and DNS records. These commands are standard for the tool's intended purpose of SEO and accessibility auditing.
  • [INDIRECT_PROMPT_INJECTION]: The skill establishes an attack surface by fetching and processing data from external domains.
  • Ingestion points: The skill fetches and reads content from remote robots.txt files and specific target $URL pages.
  • Boundary markers: None are explicitly defined in the provided command templates; the skill relies on the agent to manage the shell execution environment.
  • Capability inventory: The skill utilizes network access (curl), DNS discovery (dig), and complex text parsing (awk, sed, grep).
  • Sanitization: No explicit validation or sanitization of input variables (e.g., $SITE, $URL) is implemented within the provided audit scripts.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 3, 2026, 05:54 AM
Security Audit — agent-trust-hub — robots-ai-crawler-audit