jimubi-bigscreen

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The prompt includes command and code templates that place API keys/tokens in command-line arguments, init calls and preview URLs (e.g., bi_utils.init_api('',''), python3 ... , ?token=), so the agent would need to emit secret values verbatim — a direct exfiltration risk.

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

• Third-party content exposure detected (high risk: 0.80). This skill explicitly instructs the agent to fetch and bind open/public third‑party APIs (e.g., the listed mock endpoints like https://api.jeecg.com/mock/... in references/api-dataset-examples.md and commands such as dataset_ops.py create-api --url "https://api.jeecg.com/mock/...") as runtime data sources that the agent will read and act on to build/drive the big‑screen, so untrusted external content can influence tool use and decisions.