Skip to content
Get packs

Lark Event Listener

Stream real-time Lark/Feishu events (messages, approvals, tasks, meetings) via the lark-cli tool

Source Download skill
Repository
larksuite/cli
Stars
17,928
Licence
MIT
Files
8 files

What it does

This skill teaches an AI agent how to subscribe to live Lark/Feishu events, such as new chat messages, reactions, approval status changes, task updates, meeting start/end, minutes generated, and whiteboard updates. The agent runs the lark-cli command-line tool, which streams each event as a line of JSON that the agent can read, filter with jq expressions, and act on. It also explains how to run these listeners safely as background subprocesses, including exit codes, readiness signals, and shutdown rules.

Great for

  • Developers building a Lark or Feishu bot that reacts to messages in real time
  • Teams automating workflows when approvals, tasks, or meetings change status
  • Engineers running long-lived or time-bounded event subscribers driven by an AI agent

Skip it if

  • Users who only need one-off reads of Lark data, since this skill is about continuous event streams and the setup is heavier than a simple API call
  • Anyone without a Lark tenant and the lark-cli tool installed and authenticated, because nothing works without them
  • Running unbounded listeners under tools that close stdin, such as nohup or systemd defaults, because the listener will exit immediately unless configured otherwise

Try asking

  • “Listen for new messages in a Lark chat for the next 10 minutes and summarize them”
  • “Show me the schema for the VC meeting-ended event and write a jq filter for the meeting topic”
  • “Consume approval status changes and write each event to a file in ./events”
  • “Stream reactions and messages concurrently in two background processes”
Step-by-step guide: how to use Lark Event ListenerInstall, first run, examples, tips and FAQs

Install in 3 steps

  1. 1

    Download the skill

    The ZIP contains the 8 original files, the MIT licence and an attribution note.

    lark-event.zip
  2. 2

    Add it to your agent

    Claude Code (all projects)

    mkdir -p ~/.claude/skills && unzip lark-event.zip -d ~/.claude/skills/

    Claude Code / Agent SDK (this project only)

    mkdir -p .claude/skills && unzip lark-event.zip -d .claude/skills/

    Claude apps (web/desktop): Settings → Capabilities → Skills → Upload skill, and choose the ZIP.

  3. 3

    Ask for the task

    No special command needed — the agent reads the skill's description and loads it when your request matches. Try one of the prompts above.

Get better results

  • Use bounded runs (--max-events or --timeout) for agent-driven tasks so the process exits cleanly and returns a clear reason code
  • Run lark-cli event schema <key> --json before writing --jq expressions, since jq_root_path determines the field paths
  • Avoid --quiet for AI use, because it hides drop warnings and readiness signals, and avoid kill -9 so server-side subscriptions are cleaned up

What's inside

  • references/lark-event-application.md
  • references/lark-event-approval.md
  • references/lark-event-im.md
  • references/lark-event-minutes.md
  • references/lark-event-task.md
  • references/lark-event-vc.md
  • references/lark-event-whiteboard.md
  • SKILL.md

Redistributed unmodified from larksuite/cli under the MIT licence. The assessment was drafted with AI from the skill's files on 2026-10-29. Skills can instruct your agent to run code — read SKILL.md and any scripts before installing.