dont-acquire-for-data

Pass

Audited by Gen Agent Trust Hub on Jul 7, 2026

Risk Level: SAFENO_CODE
Full Analysis
  • [SAFE]: The skill contains no executable scripts, binaries, or shell commands. All content is strictly instructional and pedagogical.
  • [PROMPT_INJECTION]: Analysis of the instructions and test prompts confirms there are no attempts to bypass safety filters, extract system prompts, or override agent constraints.
  • [DATA_EXFILTRATION]: No hardcoded secrets, API keys, or access to sensitive local file paths (e.g., .ssh, .env) were detected.
  • [EXTERNAL_DOWNLOADS]: The skill does not perform any network operations, such as curl or wget, and contains no external URLs for code or package retrieval.
  • [REMOTE_CODE_EXECUTION]: There is no mechanism for downloading or executing code from remote sources; the skill is entirely static content.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 7, 2026, 04:04 AM
Security Audit — agent-trust-hub — dont-acquire-for-data