php-crlf-audit
Pass
Audited by Gen Agent Trust Hub on Mar 25, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill defines a specialized security auditing workflow for PHP applications. Its instructions are purely functional and consistent with its stated purpose of vulnerability discovery and documentation.
- [COMMAND_EXECUTION]: The skill instructs the agent to perform file system operations, specifically reading source code and writing markdown audit reports to a local path (
{output_path}/vuln_audit/). This is a legitimate requirement for a code auditing tool. - [DATA_EXFILTRATION]: There are no indicators of network operations, external connections, or attempts to transmit sensitive data outside the local environment.
- [PROMPT_INJECTION]: The skill does not contain instructions that attempt to override safety filters, extract system prompts, or manipulate the agent's core behavior beyond the scope of the auditing task.
Audit Metadata