c-network

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). It includes examples that embed credentials verbatim in commands (e.g., Authorization:"Bearer token123" and http -a user:password), which instructs the agent to include secret values directly in generated output and is high-risk.

CRITICAL E005: Suspicious download URL detected in skill instructions.

• Suspicious download URL detected (high risk: 0.70). cloudflare-dns.com/dns-query is a legitimate Cloudflare DNS-over-HTTPS endpoint (not a downloader) but https://example.com/file.zip is a direct .zip download from a non‑trusted/placeholder domain and direct archives from unknown sources are commonly used to distribute malware, so the overall set contains a potentially suspicious download link.

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 0.70). The SKILL.md explicitly instructs use of httpie to GET/POST/download arbitrary URLs (e.g., "http GET api.example.com/users", "http --download https://example.com/file.zip") and doggo to query public nameservers (including DoH), which clearly involves fetching and ingesting untrusted public web content that the agent would read and act on as part of its workflow.