agent-cloudflare

Pass

Audited by Gen Agent Trust Hub on Jul 16, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill implements robust security practices for credential management. It explicitly instructs the agent to never request or accept API tokens in the chat interface and utilizes a secure --form flag for local credential input into an OS dialog, ensuring secrets are handled outside the agent context.
  • [SAFE]: Potentially destructive or state-changing operations, such as DNS modifications or cache purges, are strictly gated behind a mandatory workflow requiring a --dry-run preview followed by explicit user confirmation.
  • [SAFE]: The execution environment is hardened by restricting the Bash tool to only allow specific, authenticated commands related to the agent-cloudflare and mockcloudflare CLI tools.
  • [SAFE]: While the skill ingests external data from the Cloudflare API (e.g., DNS records, logs), it does so within a well-defined operational framework that prioritizes read-only triage and guarded mutations, minimizing the risk of indirect prompt injection.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 16, 2026, 01:20 PM
Security Audit — agent-trust-hub — agent-cloudflare