skills/twilio/ai/twilio-studio-flows/Gen Agent Trust Hub

twilio-studio-flows

Pass

Audited by Gen Agent Trust Hub on Jul 23, 2026

Risk Level: SAFE
Full Analysis
  • [PROMPT_INJECTION]: No prompt injection patterns were detected. The instructions are focused on providing technical guidance for using the Twilio Studio API.
  • [DATA_EXFILTRATION]: No data exfiltration attempts were found. The skill correctly instructs users to use environment variables for sensitive credentials (TWILIO_ACCOUNT_SID, TWILIO_AUTH_TOKEN) and avoids any unauthorized network requests.
  • [EXTERNAL_DOWNLOADS]: The skill references the official 'twilio' package for both Python and Node.js, which are standard, well-known libraries for interacting with Twilio services.
  • [REMOTE_CODE_EXECUTION]: No remote code execution patterns or unsafe dynamic execution methods were identified. The provided code snippets demonstrate legitimate use of the Twilio REST API.
  • [INDIRECT_PROMPT_INJECTION]: The skill proactively identifies the risk of indirect prompt injection. It includes a specific warning in 'SKILL.md' and 'references/widgets.md' advising developers to treat inbound flow content (like SMS bodies or speech transcripts) as untrusted and to avoid interpolating such data into system prompts or commands.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 23, 2026, 06:30 PM
Security Audit — agent-trust-hub — twilio-studio-flows