linkfox-maidalv-product-tro-detection

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

• Third-party content exposure detected (low risk: 0.10). The runtime workflow only sends user-provided fields (e.g., mainProductImage, optional referenceText/description/ipKeywords) directly to the first-party /maidalv/checkApiFlash API; it does not browse/search or read any outsider-authored free-text sources on its own.

HIGH W008: Secret detected in skill content (API keys, tokens, passwords).

• Secret detected (high risk: 1.00). I scanned the skill files for hardcoded, high-entropy credentials and found one literal that appears to be a fixed, encoded UID/header used in the login flow. The value is a base64url-like string assigned to _LOGIN_FIXED_UID with an inline comment "登录链路固定 uid header（实测可用）" (fixed uid header for login, proven usable), which indicates it is a functional token rather than a documentation placeholder.

I did not flag environment variable names, placeholder text (e.g., "your-key-here", $LINKFOXAGENT_API_KEY), simple example passwords, or other instructional values — those are documentation placeholders or setup instructions and per the rules must be ignored.