lark-note

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The prompt requires embedding note/document tokens and IDs directly into CLI commands (e.g., --doc <note_doc_token>, --note-id <note_id>), which forces the LLM to output those token/identifier values verbatim and risks secret exfiltration.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.70). The SKILL explicitly mandates at runtime to "MUST 先用 Read 工具读取 ../lark-vc/references/vc-domain-boundaries.md", meaning the skill will fetch and rely on that document to control routing/decisions, so ../lark-vc/references/vc-domain-boundaries.md is a runtime external dependency that can directly control agent behavior.