audit-agentic-cli
Pass
Audited by Gen Agent Trust Hub on May 19, 2026
Risk Level: SAFE
Full Analysis
- [COMMAND_EXECUTION]: The skill includes shell scripts (
scripts/audit-cli-help.shandscripts/diff-cli-help.sh) designed to automate the inspection of CLI tools. These scripts execute user-specified binaries with flags like--helpand--versionto capture documentation. This is legitimate, documented behavior intended for the skill's primary purpose of auditing external tools. - [PROMPT_INJECTION]: The audit workflow involves the agent reading and processing output from external CLI tools. This creates a potential surface for indirect prompt injection if an audited tool's help text contains instructions designed to manipulate the agent. This is an inherent risk of the auditing task rather than a vulnerability in the skill itself, and is mitigated by the agent's standard safety guidelines.
- [SAFE]: The skill provides comprehensive security-focused documentation, such as the 'Secret Handling Rules' in
references/discovery-and-auth.md, which explicitly instructs against accepting secrets as command-line arguments and recommends secure patterns like stdin or credential helpers.
Audit Metadata