fw-ai-app-dev

Pass

Audited by Gen Agent Trust Hub on Apr 25, 2026

Risk Level: SAFE
Full Analysis
  • [PROMPT_INJECTION]: No malicious override instructions or safety bypasses detected. The skill contains strictly task-oriented instructions aligned with Freshworks Platform development.
  • [CREDENTIALS_UNSAFE]: The skill actively enforces safe credential management. It provides templates for secure installation parameters (iparams) and explicitly forbids hardcoding secrets in any code or metadata.
  • [DATA_EXFILTRATION]: No suspicious network activity or sensitive file access patterns detected. All external integrations are routed through the official Freshworks $request.invokeTemplate method, which is the platform's secure standard.
  • [REMOTE_CODE_EXECUTION]: No patterns for downloading and executing untrusted scripts were found. The skill promotes the use of standard, versioned npm packages registered in the platform manifest.
  • [OBFUSCATION]: All instructions and code templates are provided in clear, readable text. No hidden payloads or encoded commands were identified.
  • [DATA_EXPOSURE]: The skill includes a helper function in its templates specifically designed to redact URLs and sensitive information from API error responses, preventing accidental exposure of internal vendor data.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 25, 2026, 04:05 AM