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
--formflag 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-runpreview followed by explicit user confirmation. - [SAFE]: The execution environment is hardened by restricting the
Bashtool to only allow specific, authenticated commands related to theagent-cloudflareandmockcloudflareCLI 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