What problem does it solve?
AI agents and developers working with Feishu (Lark) need a reliable way to handle OAuth Device Flow login, token lifecycle, multi-app profiles, and discovery of the 2500+ OpenAPI endpoints before they can run any business operation. This Skill provides the platform foundation: authentication, diagnostics, schema browsing, raw API passthrough, global search, and directory queries.
Core Features & Use Cases
- Authentication & Token Management: Device Flow login with a two-step non-blocking mode for AI agents, scope prechecks via
auth check, token export for curl/Python, and multi-app profile isolation.
- OpenAPI Schema + Raw API Passthrough: Query local/remote OpenAPI catalogs for paths, parameters, and scopes, then call any endpoint with
feishu-cli api including dry-run, pagination, jq filtering, and error-code translation.
- Global Search & Directory: Search docs, messages (with enrichment and p2p chat support), and apps, plus read-only user and department lookups.
- Use Case: An AI agent needs to call an unwrapped Feishu endpoint. It runs
schema to find the path and required scope, prechecks with auth check --scope, performs a two-step Device Flow login, then executes the call with api --as user --dry-run before sending the real request.
Quick Start
Ask the agent to check Feishu login status and search docs for a keyword, for example: "Check my Feishu auth scopes and search docs for 'quarterly report' using feishu-cli."
