skills/efoo-team/skills/ask/Gen Agent Trust Hub

ask

Pass

Audited by Gen Agent Trust Hub on Jul 13, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill explicitly defines itself as a read-only analysis tool. It includes strict constraints prohibiting the editing, creation, or deletion of files, effectively limiting the agent's actions to data inspection and analysis.
  • [PROMPT_INJECTION]: The instructions contain a clause that prioritizes user-provided arguments over internal logic. While this can sometimes be a vector for instruction override, the skill includes a specific safety guardrail that requires the agent to halt and seek confirmation if a user request violates the core 'read-only' constraint.
  • [COMMAND_EXECUTION]: The skill limits its capabilities to directory structure confirmation, file reading, and pattern searching. There is no evidence of shell execution, privilege escalation, or persistence mechanisms.
  • [DATA_EXFILTRATION]: No network operations, external downloads, or patterns of sensitive data harvesting (such as accessing credential files) were identified.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 13, 2026, 11:19 AM
Security Audit — agent-trust-hub — ask