agents-sdk

Pass

Audited by Gen Agent Trust Hub on Jun 13, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill serves exclusively as a documentation and reference guide for the Cloudflare Agents SDK. It contains no executable logic or suspicious instructions.
  • [EXTERNAL_DOWNLOADS]: Recommends the installation of official and well-known software packages from public registries including @cloudflare/ai-chat, zod, and postal-mime. All technical documentation links lead to the trusted developers.cloudflare.com domain.
  • [COMMAND_EXECUTION]: Provides standard developer instructions for CLI-based setup and project management, such as npm install and wrangler configuration tasks.
  • [REMOTE_CODE_EXECUTION]: Documents experimental features like Codemode and Browser tools that allow for dynamic code execution. The documentation explicitly prescribes the use of isolated sandboxes (Worker sandboxes and fresh browser sessions) to ensure security and prevent unauthorized access.
  • [SAFE]: Analysis of potential indirect prompt injection vectors (such as email and webhook processing) confirms that the documentation includes best practices for security, such as HMAC-SHA256 signature verification and the use of scoped environment variables for sensitive operations.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 13, 2026, 05:06 PM