chronology

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). Yes — the skill instructs the agent to read and extract textual content (emails, docs, etc.) and write those extractions verbatim into the chronology (with source annotations) so any API keys, tokens, passwords or other secrets present in the files would be reproduced in the output.

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

• Third-party content exposure detected (high risk: 0.80). The SKILL.md explicitly states the agent will annotate and incorporate events "通过联网搜索找到的公开记录事件" and uses labels like "[联网检索——需复核]" (Sources annotation section), which shows the skill can fetch and read public/online third‑party content as part of its required workflow and those items are used to build and influence the chronology outputs.