1688-skills

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The skill explicitly instructs the agent to ask the user for their AK and to inject that AK verbatim into commands, env vars, and configuration files (e.g., ALI_1688_AK=xxx python3 ... and writing ALI_1688_AK: "your_ak_here"), which requires the LLM to handle and output secret values directly.

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

• Third-party content exposure detected (high risk: 0.90). The skill calls the external 1688 API (e.g., scripts/api.py -> POST https://ainextpre.1688.com/1688claw/skill/searchoffer) and search.py/SKILL.md explicitly instruct the Agent to read and analyze products[].stats returned from that public third‑party source and then make publishing/selection decisions (and trigger publish.py), so untrusted third‑party content is ingested and can materially influence tool usage.