check-context-communication
Pass
Audited by Gen Agent Trust Hub on Mar 17, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is entirely documentation-based, providing instructional examples and search patterns for code analysis. It does not contain any executable logic, scripts, or binary files.
- [PROMPT_INJECTION]: No evidence of instructions designed to bypass safety filters or override agent behavior. The language used is purely for instructional auditing purposes.
- [DATA_EXFILTRATION]: No access to sensitive files or network operations detected. While grep patterns search for external service names like Stripe or AWS, these are used to identify architectural coupling, not to handle credentials or perform data exfiltration.
- [COMMAND_EXECUTION]: The provided grep patterns are intended for standard static analysis of source code. They do not utilize shell injection, piping to interpreters, or other dangerous command patterns.
- [REMOTE_CODE_EXECUTION]: There are no patterns indicating the download or execution of remote scripts or the installation of third-party packages.
Audit Metadata