prompt-engineering

Pass

Audited by Gen Agent Trust Hub on Apr 15, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill is purely informational and consists of markdown documentation. It does not contain any executable scripts or configuration that could be misused.
  • [PROMPT_INJECTION]: No attempts to bypass safety filters or override agent behavior were detected. The 'System Prompt Design' section provides legitimate guidance on structuring instructions rather than malicious injection patterns.
  • [REMOTE_CODE_EXECUTION]: There are no external dependencies, package installations, or remote script downloads.
  • [DATA_EXFILTRATION]: The skill does not access sensitive local files or perform any network requests.
  • [COMMAND_EXECUTION]: No shell commands or system-level operations are included in the instructions or examples.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 15, 2026, 06:10 PM
Security Audit — agent-trust-hub — prompt-engineering