skills/xyaz1313/xyskill/xy-traffic/Gen Agent Trust Hub

xy-traffic

Pass

Audited by Gen Agent Trust Hub on Sep 18, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructions direct the agent to execute a local Python script scripts/atoms-search.py to search for specialized knowledge snippets.\n- [EXTERNAL_DOWNLOADS]: The atoms-search.py script makes network calls to api.xyskill.xyz to retrieve data, which is a non-whitelisted domain though likely associated with the vendor's ecosystem.\n- [INDIRECT_PROMPT_INJECTION]: The skill has an indirect prompt injection surface as it processes untrusted user queries through its search tools.\n
  • Ingestion points: User input provided during Phase 1 and Phase 2 of the diagnostic workflow in SKILL.md.\n
  • Boundary markers: Includes instructions defining strict role boundaries and hard constraints against bypassing platform rules.\n
  • Capability inventory: The skill utilizes a bundled script capable of local file system access and making network requests.\n
  • Sanitization: The retrieval script performs tokenization and regex-based searching but does not include advanced sanitization of user-provided query strings before they are transmitted externally.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 18, 2026, 05:22 AM
Security Audit — agent-trust-hub — xy-traffic