retellai-local-dev-loop

Pass

Audited by Gen Agent Trust Hub on Sep 9, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
  • [SAFE]: The skill provides instructional content for voice agent development and follows security best practices by advising the use of environment variables (RETELL_API_KEY) and local configuration files (.env.local) to manage credentials.\n- [EXTERNAL_DOWNLOADS]: The development setup instructions include standard installation of dependencies from the npm registry, including the official Retell SDK and common testing tools.\n- [INDIRECT_PROMPT_INJECTION]: The skill involves processing caller intents which constitutes an ingestion point for untrusted external data. While the documentation focuses on local testing with synthetic intents, this represents a known surface for indirect prompt injection in the resulting application.\n
  • Ingestion points: Synthetic caller intents (referenced in SKILL.md examples).\n
  • Boundary markers: None identified in the provided boilerplate.\n
  • Capability inventory: The skill uses the Retell SDK and has access to shell and file-write tools as defined in its platform configuration.\n
  • Sanitization: No specific sanitization or filtering logic is provided in the example code snippets.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 9, 2026, 03:49 AM
Security Audit — agent-trust-hub — retellai-local-dev-loop