cloudflare-dns

CRITICAL E004: Prompt injection detected in skill instructions.

• Potential prompt injection detected (medium risk: 0.30). The skill's purpose is ordinary Cloudflare DNS management, and the one risk is that it downloads an unverified third-party binary from a personal GitHub release into /usr/local/bin and runs it with the user's API token, but nothing shows the author intends to attack the user, so this is risky rather than malicious.

CRITICAL E005: Suspicious download URL detected in skill instructions.

• Suspicious download URL detected (medium risk: 0.30). The skill directs the agent to download an unverified, prebuilt binary from an unofficial personal GitHub account (wsvn53/flarectl) via wget, install it to /usr/local/bin, and mark it executable without any checksum or signature check, which is a concrete supply-chain risk, but nothing in the skill shows intentional malware delivery such as concealment, decoys, or a mismatch between the URL and the advertised software.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.70). The setup block downloads and installs an executable from a personal GitHub repository (wsvn53) with no checksum or signature verification, which is an unofficial, unverified source, so the tier is high.