nextjs-bisect-regression
Pass
Audited by Gen Agent Trust Hub on Jul 27, 2026
Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
- Execution of Reproduction Scripts: The workflow involves executing external reproduction code to identify pass/fail signals for specific software versions. To manage potential risks associated with running arbitrary code, the skill explicitly mandates execution within a disposable, least-privilege environment that is isolated from sensitive host data and credentials.
- Handling of External Data: The skill processes information from external sources like issue reports, repository content, and web pages. It incorporates security best practices by instructing the agent to treat all such information as untrusted data, minimizing the potential for indirect prompt injection or unintended behavior based on external content.
Audit Metadata