origin-ip-discovery

Pass

Audited by Gen Agent Trust Hub on Jul 31, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: Executes shell commands and Python snippets via the terminal to perform reconnaissance tasks, such as calculating MD5 hashes and querying DNS records.\n- [EXTERNAL_DOWNLOADS]: Fetches infrastructure data from several well-known security intelligence platforms, including SecurityTrails, AlienVault OTX, crt.sh, HackerTarget, and BuiltWith.\n- [DATA_EXFILTRATION]: Performs outbound network queries to retrieve infrastructure information for user-specified domains; no sensitive local credentials or files are exfiltrated.\n- [PROMPT_INJECTION]: The skill processes untrusted data from target websites and external APIs, which represents an indirect prompt injection attack surface.\n
  • Ingestion points: Target website HTML source, favicon files, and JSON responses from public security APIs (SKILL.md).\n
  • Boundary markers: None present; instructions do not provide explicit delimiters to separate untrusted external content from agent instructions.\n
  • Capability inventory: Terminal execution of networking tools (curl, dig) and script interpreters (python3), file system access for temporary storage, and network connectivity.\n
  • Sanitization: Basic filtering is applied using tools like grep and jq to isolate specific fields, though this does not fully eliminate the risk of natural language instruction poisoning.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 31, 2026, 04:47 PM
Security Audit — agent-trust-hub — origin-ip-discovery