blink-fetch

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The prompt includes an explicit example passing an Authorization bearer token as a command-line header (e.g., --header "Authorization: Bearer sk-xxx"), which instructs embedding secrets verbatim in generated commands and therefore poses a direct exfiltration risk.

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

• Third-party content exposure detected (high risk: 0.85). This skill performs runtime HTTP GET/POST/etc to a user-supplied arbitrary URL (blink fetch <url>), so the fetched webpage/body text from public web content can be ingested into the agent’s LLM context (indirect prompt injection risk via untrusted remote content).