web-research-subagent
Pass
Audited by Gen Agent Trust Hub on Mar 18, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [SAFE]: No malicious patterns or security risks were identified in the prompt templates or documentation files. The instructions actually promote security best practices by requiring citations and preventing context pollution.
- [NO_CODE]: This skill consists exclusively of Markdown and YAML documentation. It does not include any executable scripts, shell commands, or external code dependencies, eliminating common attack vectors such as remote code execution or privilege escalation.
Audit Metadata