e2e-test
Pass
Audited by Gen Agent Trust Hub on Apr 10, 2026
Risk Level: SAFE
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes common shell commands such as
grep,ls, andheadto identify the testing framework (e.g., Playwright, Cypress, Selenium) and read project-specific documentation. These operations are restricted to the local file system and represent standard development workflow behaviors. - [DATA_EXFILTRATION]: The skill accesses project configuration files (e.g.,
package.json,docs/project-config.json) and documentation. No network-based exfiltration patterns or access to sensitive user credentials (such as SSH keys or cloud provider tokens) were found. - [PROMPT_INJECTION]: There are no instructions aimed at overriding the agent's system prompt, bypassing safety guardrails, or extracting internal instructions.
- [REMOTE_CODE_EXECUTION]: No patterns involving the download and execution of remote scripts or unverified third-party packages were detected.
Audit Metadata