hreng-perf-diagnose
Pass
Audited by Gen Agent Trust Hub on Feb 28, 2026
Risk Level: SAFE
Full Analysis
- [COMMAND_EXECUTION]: The skill includes a shell script and a Python script used to validate the formatting of generated performance reports. These scripts perform localized file reading and pattern matching, representing a standard and safe implementation of internal validation logic.
- [SAFE]: The templates reference a JSON schema on the author's domain (clous.app). This is a routine reference for data structure validation and does not involve the execution of remote code or the exfiltration of sensitive data.
- [PROMPT_INJECTION]: While the skill is designed to ingest qualitative signals such as peer feedback, it does not possess capabilities (like network requests or shell execution based on that feedback) that could be exploited via indirect prompt injection. The analysis process is strictly for generating human-readable reports.
Audit Metadata