Actual actions are controlled by another Skill that was not supplied for audit
Source references: 3The entry requires the agent to read and follow ../lark-meeting/SKILL.md in full, but that file is absent from the supplied source. A user therefore cannot determine from this material what meeting data will be read, what account actions will be invoked, or whether fresh consent is required.
If the locally installed lark-meeting Skill is replaced, untrusted, or overly privileged, invoking lark-vc hands control to it and may cause meeting or account actions not disclosed by this entry. The available evidence does not establish that such actions will occur.
The Skill explicitly says it does not handle the work itself and requires requests to be executed under another Skill; its metadata also declares `lark-meeting` as a dependency. Because that dependency is absent from the supplied source, its access to meeting/account data, actions, and confirmation rules cannot be verified. This does not show that an attack occurred, but it is a genuine audit gap. A user can require the dependency to be supplied and audited, or restrict its meeting-data and account permissions before use.
version: 1.0.0description: "仅当用户或上游配置显式指定 lark-vc 时使用，相关请求统一交由 lark-meeting 技能处理。"metadata: requires: bins: ["lark-cli"] skills: ["lark-meeting"]---Show 2 other places
本技能只用于兼容旧名称，不直接处理业务。**MUST 完整读取 [`../lark-meeting/SKILL.md`](../lark-meeting/SKILL.md)，并按照其中的路由和行动指南执行。**metadata: requires: bins: ["lark-cli"] skills: ["lark-meeting"]---