quark-search

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The skill asks the agent to request share extraction codes (passcodes) when a share is password-protected and to include them in the validation POST body (e.g., the curl -d "passcode": "..."), which requires handling and outputting secret values verbatim.

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

• Third-party content exposure detected (high risk: 0.90). The SKILL.md workflow explicitly instructs the agent to query the public PanSou search API (https://s.panhunt.com/api/search) and consume public cloud-share pages (e.g., https://pan.quark.cn/s/...) — untrusted, user-generated content — and to parse those results to validate links and decide which resources to save via the local Quark app, so third-party content can materially influence agent actions and enable indirect prompt injection.