AGENT SKILL

meeting-sdk/linux

Zoom Meeting SDK for Linux - C++ headless meeting bots with raw audio/video access, transcription, recording, and AI integration for server-side automation

License 许可
Apache-2.0
最近更新
2026-11-11
Stars
★ 3.5w
Forks
3.9k

01 前言 Front Matter

name
meeting-sdk/linux
triggers
[ "linux meeting bot", "headless zoom bot", "meeting sdk linux", "zoom raw recording linux", "meeting transcription bot", "zoom docker bot", "pulseaudio zoom", "meeting sdk raw data" ]
description
Zoom Meeting SDK for Linux - C++ headless meeting bots with raw audio/video access, transcription, recording, and AI integration for server-side automation
user-invocable
false

02 资源文件 5

03 SKILL.md 正文

Zoom Meeting SDK - Linux Development

Expert guidance for building headless meeting bots with the Zoom Meeting SDK on Linux. This SDK enables server-side meeting participation, raw media capture, transcription, and AI-powered meeting automation.

How to Build a Meeting Bot That Automatically Joins and Records

Use this skill when the requirement is:

  • visible bot joins a real Zoom meeting
  • the bot records raw media itself
  • or the bot triggers a Zoom-managed cloud-recording workflow after join

Skill chain:

  • primary: meeting-sdk/linux
  • add zoom-rest-api for OBF/ZAK lookup, scheduling, or cloud-recording settings
  • add zoom-webhooks when post-meeting cloud recording retrieval is required

Minimal raw-recording flow:

JoinParam join_param;
join_param.userType = SDK_UT_WITHOUT_LOGIN;
auto& params = join_param.param.withoutloginuserJoin;
params.meetingNumber = meeting_number;
params.userName = "Recording Bot";
params.psw = meeting_password.c_str();
params.app_privilege_token = obf_token.c_str();

SDKError join_err = meeting_service->Join(join_param);
if (join_err != SDKERR_SUCCESS) {
    throw std::runtime_error("join_failed");
}

// In MEETING_STATUS_INMEETING callback:
auto* record_ctrl = meeting_service->GetMeetingRecordingController();
if (!record_ctrl) {
    throw std::runtime_error("recording_controller_unavailable");
}

if (record_ctrl->CanStartRawRecording() != SDKERR_SUCCESS) {
    throw std::runtime_error("raw_recording_not_permitted");
}

SDKError record_err = record_ctrl->StartRawRecording();
if (record_err != SDKERR_SUCCESS) {
    throw std::runtime_error("start_raw_recording_failed");
}

GetAudioRawdataHelper()->subscribe(new MyAudioDelegate());

Use raw recording when the bot must own PCM/YUV media or feed an AI pipeline directly.
Use cloud recording + webhooks when the requirement is Zoom-managed MP4/M4A/transcript assets after the meeting.

Official Documentation: https://developers.zoom.us/docs/meeting-sdk/linux/
API Reference: https://marketplacefront.zoom.us/sdk/meeting/linux/
Sample Repository (Raw Recording): https://github.com/zoom/meetingsdk-linux-raw-recording-sample
Sample Repository (Headless): https://github.com/zoom/meetingsdk-headless-linux-sample

Quick Links

New to Meeting SDK Linux? Follow this path:

  1. linux.md - Quick start guide with complete workflow
  2. concepts/high-level-scenarios.md - Production bot architectures
  3. meeting-sdk-bot.md - Resilient bot with retry logic
  4. references/linux-reference.md - Dependencies, Docker, CMake

Common Use Cases:

Having issues?

Routing Rule for Bots

If the user asks to build a bot that automatically joins a Zoom meeting and records it, start with meeting-sdk-bot.md.

  • Use Meeting SDK Linux for the visible participant, join flow, and raw recording control.
  • Chain zoom-rest-api when the bot must fetch OBF/ZAK tokens, schedule meetings, or enable account-side recording settings.
  • Chain zoom-webhooks when the requirement is Zoom cloud recording retrieval after meeting end.

SDK Overview

The Zoom Meeting SDK for Linux is a C++ library optimized for headless server environments:

  • Headless Operation: No GUI required, perfect for Docker/cloud
  • Raw Data Access: YUV420 video, PCM audio at 32kHz
  • GLib Event Loop: Async event handling for callbacks
  • Docker-Ready: Pre-configured Dockerfiles for CentOS/Ubuntu
  • PulseAudio Integration: Virtual audio devices for headless environments

Key Differences from Video SDK

Feature Meeting SDK (Linux) Video SDK
Primary Use Join existing meetings as bot Host custom video sessions
Visibility Visible participant Session participant
UI Headless (no UI) Optional custom UI
Authentication JWT + OBF/ZAK for external meetings JWT only
Recording Control StartRawRecording() required Direct raw data access
Platform Linux only Windows, macOS, iOS, Android

Prerequisites

System Requirements

  • OS: Ubuntu 22+, CentOS 8/9, Oracle Linux 8
  • Architecture: x86_64
  • Compiler: gcc/g++ with C++11 support
  • Build Tools: cmake 3.16+

Development Dependencies

# Ubuntu
apt-get install -y build-essential cmake \
    libx11-xcb1 libxcb-xfixes0 libxcb-shape0 libxcb-shm0 \
    libxcb-randr0 libxcb-image0 libxcb-keysyms1 libxcb-xtest0 \
    libglib2.0-dev libcurl4-openssl-dev pulseaudio

# CentOS
yum install -y cmake gcc gcc-c++ \
    libxcb-devel xcb-util-image xcb-util-keysyms \
    glib2-devel libcurl-devel pulseaudio

Required Credentials

  1. Zoom Meeting SDK App (Client ID & Secret) → Create at Marketplace
  2. JWT Token → Generate from Client ID/Secret
  3. For External Meetings: OBF token OR ZAK token → Get via REST API
  4. For Raw Recording: Meeting Recording Token (optional) → Get via API

Quick Start

1. Download & Extract SDK

# Download from https://marketplace.zoom.us/
tar xzf zoom-meeting-sdk-linux_x86_64-{version}.tar

# Organize files
mkdir -p demo/include/h demo/lib/zoom_meeting_sdk
cp -r h/* demo/include/h/
cp lib*.so demo/lib/zoom_meeting_sdk/
cp -r qt_libs demo/lib/zoom_meeting_sdk/
cp translation.json demo/lib/zoom_meeting_sdk/json/

# Create required symlink
cd demo/lib/zoom_meeting_sdk && ln -s libmeetingsdk.so libmeetingsdk.so.1

2. Initialize & Auth

#include "zoom_sdk.h"

USING_ZOOM_SDK_NAMESPACE

// Initialize SDK
InitParam init_params;
init_params.strWebDomain = "https://zoom.us";
init_params.enableLogByDefault = true;
init_params.rawdataOpts.audioRawDataMemoryMode = ZoomSDKRawDataMemoryModeHeap;
InitSDK(init_params);

// Authenticate with JWT
AuthContext auth_ctx;
auth_ctx.jwt_token = your_jwt_token;
CreateAuthService(&auth_service);
auth_service->SDKAuth(auth_ctx);

3. Join Meeting

// In onAuthenticationReturn callback
void onAuthenticationReturn(AuthResult ret) {
    if (ret == AUTHRET_SUCCESS) {
        JoinParam join_param;
        join_param.userType = SDK_UT_WITHOUT_LOGIN;
        
        auto& params = join_param.param.withoutloginuserJoin;
        params.meetingNumber = 1234567890;
        params.userName = "Bot";
        params.psw = "password";
        params.isVideoOff = true;
        params.isAudioOff = false;
        
        meeting_service->Join(join_param);
    }
}

4. Access Raw Data

// In onMeetingStatusChanged callback
void onMeetingStatusChanged(MeetingStatus status, int iResult) {
    if (status == MEETING_STATUS_INMEETING) {
        auto* record_ctrl = meeting_service->GetMeetingRecordingController();
        
        // Start raw recording (enables raw data access)
        if (record_ctrl->CanStartRawRecording() == SDKERR_SUCCESS) {
            record_ctrl->StartRawRecording();
            
            // Subscribe to audio
            auto* audio_helper = GetAudioRawdataHelper();
            audio_helper->subscribe(new MyAudioDelegate());
            
            // Subscribe to video
            IZoomSDKRenderer* video_renderer;
            createRenderer(&video_renderer, new MyVideoDelegate());
            video_renderer->setRawDataResolution(ZoomSDKResolution_720P);
            video_renderer->subscribe(user_id, RAW_DATA_TYPE_VIDEO);
        }
    }
}

Key Features

Feature Description
Headless Operation No GUI, perfect for Docker/server deployments
Raw Audio (PCM) Capture mixed or per-user audio at 32kHz
Raw Video (YUV420) Capture video frames in contiguous planar format
GLib Event Loop Async callback handling
Docker Support Pre-built Dockerfiles for CentOS/Ubuntu
PulseAudio Virtual Devices Audio in headless environments
Breakout Rooms Programmatic breakout room management
Chat Send/receive in-meeting chat
Recording Control Local, cloud, and raw recording

Critical Gotchas & Best Practices

⚠️ CRITICAL: PulseAudio for Docker/Headless

The #1 issue for raw audio in Docker:

Raw audio requires PulseAudio and a config file, even in headless environments.

Solution:

# Install PulseAudio
apt-get install -y pulseaudio pulseaudio-utils

# Create config file
mkdir -p ~/.config
cat > ~/.config/zoomus.conf << EOF
[General]
system.audio.type=default
EOF

# Start PulseAudio with virtual devices
pulseaudio --start --exit-idle-time=-1
pactl load-module module-null-sink sink_name=virtual_speaker
pactl load-module module-null-sink sink_name=virtual_mic

See: references/linux-reference.md#pulseaudio-setup

Raw Recording Permission Required

Unlike Video SDK, Meeting SDK requires explicit permission to access raw data:

// MUST call StartRawRecording() first
auto* record_ctrl = meeting_service->GetMeetingRecordingController();

SDKError can_record = record_ctrl->CanStartRawRecording();
if (can_record == SDKERR_SUCCESS) {
    record_ctrl->StartRawRecording();
    // NOW you can subscribe to audio/video
} else {
    // Need: host/co-host status OR recording token
}

Ways to get permission:

  1. Bot is host/co-host
  2. Host grants recording permission
  3. Use recording_token parameter (get via REST API)
  4. Use app_privilege_token (OBF) when joining

GLib Main Loop Required

SDK callbacks execute via GLib event loop:

#include <glib.h>

// Setup main loop
GMainLoop* loop = g_main_loop_new(NULL, FALSE);

// Add timeout for periodic tasks
g_timeout_add_seconds(10, check_meeting_status, NULL);

// Run loop (blocks until quit)
g_main_loop_run(loop);

Without GLib loop: Callbacks never fire, join hangs indefinitely.

Heap Memory Mode Recommended

Always use heap mode for raw data to avoid stack overflow with large frames:

init_params.rawdataOpts.videoRawDataMemoryMode = ZoomSDKRawDataMemoryModeHeap;
init_params.rawdataOpts.shareRawDataMemoryMode = ZoomSDKRawDataMemoryModeHeap;
init_params.rawdataOpts.audioRawDataMemoryMode = ZoomSDKRawDataMemoryModeHeap;

Thread Safety

SDK callbacks execute on SDK threads:

  • Don't perform heavy operations in callbacks
  • Don't call CleanUPSDK() from within callbacks
  • Use thread-safe queues for data passing
  • Use mutexes for shared state

Production Architectures

Transcription Bot

See: concepts/high-level-scenarios.md#scenario-1

Join Meeting → StartRawRecording → Subscribe Audio → 
Stream to AssemblyAI/Whisper → Generate Real-time Transcript

Recording Bot

See: concepts/high-level-scenarios.md#scenario-2

Join Meeting → StartRawRecording → Subscribe Audio+Video → 
Write YUV+PCM → FFmpeg Merge → Upload to Storage

AI Meeting Assistant

See: concepts/high-level-scenarios.md#scenario-3

Join Meeting → Real-time Transcription → AI Analysis → 
Extract Action Items → Generate Summary

Sample Applications

Official Repositories:

Sample Description Link
Raw Recording Sample Traditional C++ approach with config.txt GitHub
Headless Sample Modern TOML-based with CLI, Docker Compose GitHub

What Each Demonstrates:

  • Raw Recording Sample: Complete raw data workflow, PulseAudio setup, Docker multi-distro support
  • Headless Sample: AssemblyAI integration, Anthropic AI, production-ready config management

Documentation Library

Core Concepts

Platform Reference

Authentication

Advanced Features

Common Issues

Issue Cause Solution
No audio in Docker Missing PulseAudio config Create ~/.config/zoomus.conf
Raw recording denied No permission Use host/co-host OR recording token
Callbacks not firing Missing GLib main loop Add g_main_loop_run()
Build errors (XCB) Missing X11 libraries Install libxcb packages
Segfault on auth OpenSSL version mismatch Install libssl1.1

Complete troubleshooting: references/linux-reference.md#troubleshooting

Resources


Need help? Start with linux.md for quick start, then explore high-level-scenarios.md for production patterns.

Operations

  • RUNBOOK.md - 5-minute preflight and debugging checklist.

04 同仓库更多 Skills

anthropics
anthropics/knowledge-work-plugins

ticket-deflector

Reads a forwarded customer email or ticket, pulls order and refund status from a payments connector (PayPal, Square, or Stripe) or Shopify, account history from the CRM, and open tickets from a support desk (Zoho Desk), drafts a tone-matched reply in the owner's writing voice, and can issue a refund through the payments connector with explicit owner approval. With Shopify connected it also runs a proactive order-triage mode that surfaces orders needing attention — unfulfilled past the promised window, payment problems, pending refunds, stuck shipments — and drafts the next action for each before the customer has to ask. Use when the user says "draft a response," "answer this customer," "where's my order," "I want a refund," "check my orders," or "anything about to blow up."

★ 3.5w
anthropics
anthropics/knowledge-work-plugins

tax-season-organizer

Prepares tax-season materials for the owner's accountant, not tax advice. US federal tax; a non-US business gets its closed-books packet instead. Two modes: (1) quarterly estimated tax from YTD net income in the ledger (MYOB, NetSuite, QuickBooks, Xero, or Zoho Books); (2) year-end 1099 prep, scanning the ledger, PayPal, and Stripe for contractors paid over USD 600 into a 1099-NEC list with missing W-9 flags. Any tax request routes first to /tax-prep, which confirms the books are closed and reconciled before running this skill. Use this skill directly only when the owner says the period's books are already closed: "books are closed, now do the 1099s," "run the quarterly estimate off the closed numbers," or "just the contractor W-9 list."

★ 3.5w
anthropics
anthropics/knowledge-work-plugins

tax-prep

Prepares tax materials as a two-link chain — month-end-prep confirms the books are closed and reconciled first, then tax-season-organizer calculates the quarterly estimated payment or builds the year-end 1099-NEC list and accountant packet from those closed numbers. Requires a ledger (MYOB, NetSuite, QuickBooks, Xero, or Zoho Books); uses Gusto, PayPal, and Stripe when connected, else CSV upload. US federal tax math; a non-US business gets the closed-books packet. Prep material for a CPA, never tax advice, nothing filed. Trigger on "quarterly taxes," "estimated tax payment," "how much should I set aside," "1099s," "1099-NEC," "W-9s," "year-end tax prep," "get my books ready for my accountant," any phrasing that suggests a tax deadline is coming, or a question about net profit or YTD income that sounds like worry about a tax bill. An owner who says the period's books are already closed routes to tax-season-organizer directly.

★ 3.5w
anthropics
anthropics/knowledge-work-plugins

speed-to-lead

Makes sure no inbound lead goes unanswered. Watches web forms, the shared inbox, and the CRM for new inquiries, qualifies each one against the owner's criteria, drafts a reply in the owner's voice with real meeting times for the owner to approve and send, routes anything hot to a human right away, and logs the whole thing to the CRM. Never sends on its own. Built to run on a schedule so overnight inquiries are drafted and sorted by morning, and works from forwarded or pasted inquiries when no connector is available. Use this whenever the owner is worried about inbound leads or response time — including phrasings like "leads are going cold," "did anyone answer that inquiry," "I keep missing form submissions," "nothing falls through the cracks," "handle my inbound," "respond to this lead," or "we lose people because nobody gets back fast enough." Reach for it for a single unanswered inquiry as readily as for a standing setup.

行政与办公人员
★ 3.5w
anthropics
anthropics/knowledge-work-plugins

social-content-engine

Runs the owner's content operation as a standing thing rather than a one-off: keeps a rolling posting calendar, learns the owner's voice, generates on-brand Canva graphics, writes captions that sound like them, repurposes one asset into every channel format, and stages posts for approval instead of publishing on its own. Pulls product photos and prices from Shopify when connected; with no connectors, drafts the calendar and asset briefs for manual posting. Use this whenever the owner talks about posting, content, or social — including "I need to post more," "we've gone quiet on Instagram," "what should we post this month," "make the content," "give me a month of posts," "write this in my voice," "repurpose this," or "set up our content calendar." Reach for it for a single post as readily as for a full quarter. One handoff: a finished, approved campaign brief goes to canva-creator; this skill owns the standing operation.

内容创作者
★ 3.5w
anthropics
anthropics/knowledge-work-plugins

smb-router

The front door to the Small Business plugin. Listens to what the owner needs right now — vague or specific — and routes them to the best skill or slash command for the moment. Also serves as a guide: explains what's available, suggests what to try next, and adapts recommendations based on stored business context. Trigger whenever the owner asks "what can you do," "help me with my business," "what should I focus on," "I don't know where to start," or any open-ended business request that doesn't clearly match a single skill.

★ 3.5w