hunt-business-logic

Fail

Audited by Gen Agent Trust Hub on Jul 31, 2026

Risk Level: HIGHPROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill processes external, untrusted data including robots.txt, sitemap.xml, and JavaScript bundles. This creates a surface for indirect prompt injection where an attacker could place instructions in these files to influence the agent behavior.
  • Ingestion points: target's robots.txt, sitemap.xml, and JavaScript assets.
  • Boundary markers: Absent.
  • Capability inventory: curl, grep.
  • Sanitization: Absent.
  • [COMMAND_EXECUTION]: Uses curl and grep to perform active probes, payment tampering simulations, and metadata extraction from targets.
  • [EXTERNAL_DOWNLOADS]: Interacts with httpbin.org, a well-known service, to verify request formatting and connectivity.
  • [REMOTE_CODE_EXECUTION]: Pipes the output of network requests to python3 -m json.tool. This uses the Python interpreter for the safe purpose of formatting JSON data rather than executing external scripts.
Recommendations
  • HIGH: Downloads and executes remote code from: https://httpbin.org/post - DO NOT USE without thorough review
Audit Metadata
Risk Level
HIGH
Analyzed
Jul 31, 2026, 04:47 PM
Security Audit — agent-trust-hub — hunt-business-logic