region-config
Pass
Audited by Gen Agent Trust Hub on Feb 27, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: No patterns of behavior override or safety bypass were identified.
- [DATA_EXPOSURE_AND_EXFILTRATION]: The skill manages marketing profiles and regional settings within the ~/.claude-marketing/ local directory. These actions are limited to the local filesystem and are required for the skill's primary functionality. No evidence of data exfiltration was found.
- [INDIRECT_PROMPT_INJECTION]: The skill ingests user input to define regional settings which are then persisted in configuration files.
- Ingestion points: User input for region name, currency, platforms, and industry regulations.
- Boundary markers: No explicit markers are used to separate untrusted inputs in the generated JSON configs.
- Capability inventory: File system write access to local brand directories.
- Sanitization: No specific input validation or sanitization is described for user-supplied configuration data before it is saved.
Audit Metadata