load-test-scenario-builder
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFE
Full Analysis
- [COMMAND_EXECUTION] (SAFE): The provided bash script executes the k6 binary to run performance tests. This is the intended functionality of the skill and does not involve risky shell patterns or unauthorized commands.
- [DATA_EXFILTRATION] (SAFE): HTTP requests in the sample scripts are directed toward example.com and a local InfluxDB instance (localhost). There is no evidence of sensitive data harvesting or exfiltration to external domains.
- [REMOTE_CODE_EXECUTION] (SAFE): The skill does not download or execute remote scripts. The k6 execution is limited to local files defined within the skill context.
- [PROMPT_INJECTION] (SAFE): The instructions and metadata are focused on technical load testing and do not contain language intended to override agent safety protocols.
Audit Metadata