app-dev
Pass
Audited by Gen Agent Trust Hub on Mar 3, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is authored by 'freshworks-developers' and provides legitimate tools and resources for the Freshworks developer ecosystem.
- [PROMPT_INJECTION]: The skill uses strong instructional language and enforcement markers (e.g., 'ZERO TOLERANCE', 'MANDATORY') to ensure the agent adheres to Platform 3.0 specifications and rejects legacy patterns. These are technical constraints for code generation and do not attempt to override agent safety protocols.
- [COMMAND_EXECUTION]: Includes a utility script ('scripts/cleanup_old_rules.sh') for removing specific rule files from the project environment. The script is narrowly scoped to specific Freshworks configuration files and serves a standard maintenance purpose.
- [REMOTE_CODE_EXECUTION]: References to external packages and UI libraries point to official registries and official vendor CDNs (such as freshdev.io and unpkg.com). These are standard dependencies for the platform.
- [DATA_EXFILTRATION]: No patterns of unauthorized data collection or external transmission were found. The skill emphasizes security by instructing the agent to use request templates and secure installation parameters for sensitive data handling.
Audit Metadata