agent-native-design
Pass
Audited by Gen Agent Trust Hub on Jul 8, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: The skill provides instructions for analyzing CLI designs and does not contain any patterns intended to bypass safety filters or override agent behavior.
- [DATA_EXFILTRATION]: No evidence of hardcoded credentials or unauthorized network operations. The skill uses environment variables as a recommended design pattern for authentication delegation, which aligns with security best practices.
- [REMOTE_CODE_EXECUTION]: No remote code execution patterns detected. The reference files include examples of shell commands and testing scripts intended for users to verify their own CLI tools, not for the agent to execute maliciously.
- [COMMAND_EXECUTION]: The skill contains a metadata validation script (
scripts/validate-metadata.sh) used for repository maintenance. This script performs benign local operations such as version comparison and file existence checks using standard utilities (rg,git,sort). - [INDIRECT_PROMPT_INJECTION]: The skill has a data ingestion surface as it processes user-provided CLI help text and API specifications for review. However, it lacks the dangerous capabilities (like automated code execution on that data) that would make it exploitable, and it provides a structured rubric for analysis.
Audit Metadata