Seminara

io.github.shivamselam由 GitHub 账号 shivamselam 发布v1.0.0更新于 Oct 26, 2026

Digital Teammate (Aura) hosting live interactive presentations, demos, and audience Q&A 24/7.

已验证Streamable HTTP可网页运行沟通与协作媒体与设计AI 与机器学习

概览

AI 生成的概览

Seminara 提供一个远程 MCP 端点,让 AI 助手创建并运行由自主智能体 Aura 主持的实时语音演示。

可以做什么
  • 根据标题、PDF 演示文稿链接和问答知识库创建实时演示会话
  • 列出并管理正在进行的演示房间
  • 为会话添加带文字和链接的行动号召
  • 由 Aura 主持交互式幻灯片并实时回答观众提问
  • 通过 WebMCP、CLI 和多语言 SDK 访问同一套演示引擎
适用场景
适合希望让助手自动搭建并主持实时演示或产品演示房间(含观众问答)的场景,例如产品演示、销售沟通会或发布活动,无需真人主持。
运行要求
使用 seminara.online 上的远程 MCP 端点,无需本地运行时。核心产品服务器需要 API 密钥,通过 Bearer Authorization 请求头传递(SDK 和 CLI 示例中使用 SEMINARA_API_KEY)。需要 Seminara 账号和网络访问。
安装前请注意
API 密钥是正式凭据(示例中形如 sk_live_...),应避免写入共享配置。会话在 Seminara 服务器上创建,你提供的演示文稿和知识库内容会发送给该第三方。创建会话可能消耗付费套餐;README 提到 Core Host 套餐为每月 199 美元、含 50 场演示。

安装

在 SourceWeft 中

  1. 打开 控制台中的 Seminara,将其添加到工作区。
  2. 为需要使用其工具的对话启用该服务。

Web executable,通过 Streamable HTTP。 远程服务在工作区中配置后即可从网页运行时运行。

交给智能体安装

把这段提示词粘贴到 Claude Code、Codex、Cursor 等智能体,或 SourceWeft 对话中。智能体会读取该服务器的安装指南,先向你展示发布者、端点和登录方式,经你同意后再把它添加到你的客户端。密钥不会写进任何命令。

阅读 https://sourceweft.com/mcp/io.github.shivamselam%2Fseminara/install.md,并按其中说明把这个 MCP 服务器接入我的客户端。修改任何配置之前,先告诉我它的发布者、端点和登录方式,等我确认。不要把 API 密钥写进命令或我会提交的文件;问我密钥应从哪里获取。

查看智能体所依据的安装指南

其他 MCP 客户端

把它添加到你客户端的 mcpServers 配置中。

{
  "mcpServers": {
    "seminara": {
      "type": "http",
      "url": "https://seminara.online/api/v1/mcp"
    }
  }
}

README

The Presentation Layer for AI Agents

Seminara is an autonomous, 24/7 WebRTC presentation host (Aura) that delivers interactive slide decks and fields live audience Q&A.

With the Seminara Agentic Suite, this core engine is now fully programmable. You or your autonomous agents can orchestrate live, voice-led presentations entirely through code using our WebMCP integration, CLI, or multi-language SDKs.


📦 Available Packages & Tools

Package / ToolLanguageDirectoryInstallation
seminara-sdkTypeScript / JavaScriptsdk/npm install seminara-sdk
seminaraPython 3.9+python/pip install seminara
seminara-cliCommand Line CLIcli/npm install -g seminara-cli
goGo 1.21+go/go get github.com/shivamselam/seminara-agentic-suite/go
[Go Reference]

🚀 Quickstarts: Time to Hello World < 5 Minutes

1. Node.js / TypeScript

bash
npm install seminara-sdk
typescript
import { Seminara } from 'seminara-sdk';
const client = new Seminara({ apiKey: process.env.SEMINARA_API_KEY });
// Instantly generate a live, AI-hosted presentation roomconst session = await client.sessions.create({  title: "Q3 Enterprise Demo & Discovery",  pdfUrl: "https://example.com/pitch-deck.pdf",  knowledgeBase: [    { title: "Pricing", content: "Core Host plan is $199/mo for 50 presentations." }  ],  ctaText: "Book Architecture Review",  ctaUrl: "https://seminara.online/book-meeting"});
console.log(`Live Session URL: ${session.liveLink}`);

2. Python

bash
pip install seminara
python
import osfrom seminara import SeminaraClient
client = SeminaraClient(api_key=os.environ.get("SEMINARA_API_KEY"))
session = client.create_session(    title="Q3 Enterprise Demo & Discovery",    is_live=True,    pdf_url="https://example.com/pitch-deck.pdf",    knowledge_base=[        {"title": "Pricing", "content": "Core Host plan is $199/mo for 50 presentations."}    ],    cta_text="Book Architecture Review",    cta_url="https://seminara.online/book-meeting")
print(f"Live Session Created: {session['liveLink']}")

3. Command Line (CLI)

bash
npm install -g seminara-cli
# Set your API keyexport SEMINARA_API_KEY="sk_live_..."
# Create a presentation room straight from your terminalseminara create --title "Q3 Product Launch"
# List active roomsseminara list

🤖 Model Context Protocol (MCP) & AI Connectors

Seminara natively exposes its presentation engine across the entire agent ecosystem:

  • Standard Well-Known Discovery: https://seminara.online/.well-known/mcp
  • Core Product MCP Server: https://seminara.online/api/v1/mcp
  • Documentation MCP Server: https://seminara.online/api/docs/mcp
  • OpenAPI 3.1 Spec (ChatGPT / Custom GPTs): https://seminara.online/openapi.json
  • Discovery Catalog: https://seminara.online/.well-known/api-catalog.json
  • Interactive Sandbox & Zero-Cost Rehearsals: https://seminara.online/sandbox
  • Cursor Directory Plugin: cursor.directory/plugins/seminara-agentic-suite

1-Click Install via Cursor Directory

Install the complete plugin package (Aura subagent, rules, skills, and MCP servers) directly from the Cursor Directory:

[Install in Cursor]

1-Click Install via Smithery CLI

Installs automatically into Claude Desktop, Cursor, Windsurf, Cline, Zed, and more:

bash
npx -y @smithery/cli install @seminara/autonomous-host

(Or specify client directly: --client claude, --client cursor, --client windsurf, --client cline)

Quick Connect Configurations

Claude Desktop (claude_desktop_config.json)
json
{  "mcpServers": {    "seminara": {      "type": "sse",      "url": "https://seminara.online/api/v1/mcp",      "headers": {        "Authorization": "Bearer sk_live_your_key_here"      }    }  }}
Cursor & Windsurf (.cursor/mcp.json or mcp_config.json)
json
{  "mcpServers": {    "seminara": {      "type": "sse",      "url": "https://seminara.online/api/v1/mcp"    }  }}
ChatGPT / OpenAI Custom GPTs (Actions)
  1. In your GPT configuration, go to Actions → Create new action.
  2. Click Import from URL and paste: https://seminara.online/openapi.json
  3. Under Authentication, select Bearer and provide your API key.
WebMCP (In-Browser Agents)

If you navigate to https://seminara.online using a WebMCP-enabled browser, your agent will automatically discover the exposed presentation tools on page load.


📖 Documentation & Resources


📄 License

MIT License. Copyright (c) 2026 Shivam Selam / Omni AI Club Private Limited.

来源:README.md,提交 4e54d67

工具

0
SourceWeft 尚未读取此服务器的工具。

版本历史

1
  1. v1.0.0最新Sep 16, 2026

评分

已满足适用检查项的 85%

该评分衡量此收录条目的完整程度,依据注册中心条目、端点与公开来源计算得出。它不是对服务器本身的评价,也不表示服务器安全或当前可用:可连通性是带日期的检查结果,评分也不包含安装量或用户评价。

  • 近 30 天内可连通必需25 分
  • 已声明认证方式必需10 分
  • 已列出工具必需15 分

    请让服务器在无凭据时响应 tools/list,或发布可在沙箱中无凭据启动的软件包。

  • 描述必需15 分
  • 图标10 分
  • 文档15 分
  • 许可证10 分