fw-ai-actions-app
Pass
Audited by Gen Agent Trust Hub on May 14, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill provides comprehensive and safe guidance for Freshworks AI Action development with no malicious code or instructions detected. All resource links target official Freshworks developer documentation or the author's official GitHub repository.
- [CREDENTIALS_UNSAFE]: Implements mandatory rules against hardcoding secrets in any app file. All templates and agent instructions enforce the use of secure installation parameters (
iparams) or OAuth for credential management. - [COMMAND_EXECUTION]: CLI tool usage is strictly limited to the Freshworks Developer Kit (FDK) for validation and local testing. The included agents (validator, scoper, implementer) have clearly defined guardrails and require user confirmation for destructive or cross-app operations.
- [REMOTE_CODE_EXECUTION]: No patterns for unverified remote code execution or dangerous package installations were found. The skill recommends well-known, legitimate dependencies such as
jsonwebtoken,strong-soap, andunderscorefor specific integration tasks. - [PROMPT_INJECTION]: The skill proactively mitigates indirect prompt injection risks by enforcing flat request schemas (reducing input surface complexity) and providing sanitization patterns (
__sanitizeMessageForClient) to prevent sensitive data leakage from third-party API responses.
Audit Metadata