Feishu CLI Web

Turn Flybook/Lark CLI into a privatizable and deployable Web smart workbench. Feishu CLI Web is based on the official lark-cli and provides natural language interaction, execution plan preview, write operation confirmation, multi-user isolation and SQLite local storage. It is suitable for teams to quickly connect flying book automation capabilities to their own Agents, internal tools or operation and maintenance platforms. Feishu's official CLI capabilities are complete, but team members usually don't want to remember commands, configuration parameters, process authorizations and debugging output. Feishu CLI Web adds a layer of Web workflows more suitable for teams on top of CLI: 1. Describing the flying book mission in natural language 2. Preview plans and orders before execution 3. Write operations need to be confirmed to reduce the risk of misoperation 4. Each user authorizes independently and does not affect each other 5. Chat records, execution records, and account information are uniformly saved in SQLite 6. Built-in common scene templates to facilitate the precipitation of team processes 7. Built-in AI PPT workflow, supports generation, modification, full preview, and uploads cloud documents with buttons or sends them to Flying Book 8. One deployment, multiple people use it: Each user has a separate account and does not affect each other. 9. Agents focusing on Flybook CLI capabilities: Super fast response speed. Description: 1. Support lazy cat OIDC login and local login. The local login account password defaults to admin/12345678 2. For the first time, you need to configure the AI model and authorize the flying book. Before performing certain operations, you need to authorize the robot rights. You can authorize it online and authorize it in the background of flying book management. If online authorization fails, you can click the "Supplementary Authorization" button to try authorization again 3. The experience and convenience of use have a lot to do with the AI model. It is recommended to use a smarter model.