convex-security
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFE
Full Analysis
- [Prompt Injection] (SAFE): No instructions were found that attempt to override agent behavior or bypass safety filters. The content is strictly instructional.
- [Data Exposure & Exfiltration] (SAFE): No hardcoded credentials or unauthorized network operations were detected. Database access snippets (ctx.db) follow recommended security patterns for the framework.
- [Remote Code Execution] (SAFE): The skill does not perform remote script execution or download unverifiable executable content. References to external documentation point to official Convex domains.
- [Obfuscation] (SAFE): The file contains clear, human-readable text and code without any hidden or encoded content.
- [Indirect Prompt Injection] (LOW): The skill describes handling user-provided data but mitigates risk by enforcing strict schema validation (args/returns validators) and robust authentication helpers.
- [Unverifiable Dependencies] (SAFE): Mentioned packages and ESLint rules are standard for the Convex ecosystem and do not involve suspicious installation patterns.
Audit Metadata