Plaud

io.github.mcp-dir由 GitHub 账号 mcp-dir 发布v0.1.0更新于 Oct 24, 2026

Your Plaud recordings in natural language: list recordings, read speaker-attributed transcripts and

已验证Streamable HTTP可网页运行效率与工作流知识与记忆

概览

AI 生成的概览

把 Plaud 录音接入 AI 助手:列出录音、读取带说话人标注的转写文本,并获取 AI 摘要与待办事项。

可以做什么
  • 按名称子串和日期范围筛选并分页列出录音
  • 读取单条录音详情,或按 ID 批量读取多条
  • 读取带时间戳和说话人归属的完整转写文本
  • 读取 AI 生成的笔记:精简摘要、待办事项和关键主题
  • 读取已授权 Plaud 用户的资料
  • 查看连接状态与工具包状态、平台版本,并提交反馈或缺陷报告
适用场景
当你希望助手回顾用 Plaud 记录的会议或语音备忘内容时适用,例如总结通话、提取待办事项或引用谁说了什么,无需手动导出文件。
运行要求
通过 HTTP 访问的远程 MCP 端点,无需本地安装包或运行时。需要一个 Plaud 账号,通过一键连接流程授权(无需粘贴 API 密钥)。部分客户端可能需要以 Authorization Bearer 请求头提供访问令牌。
安装前请注意
录音、转写文本和笔记会发送到你选择的 LLM 宿主,README 将其列为不受提供方控制的次级处理者,并建议使用可选择不用于训练的套餐。Plaud 相关工具据称只读,但服务器还提供账号、市场与反馈类工具,且缺陷报告工具会发送最近的对话消息。该托管服务器为专有软件。

安装

在 SourceWeft 中

  1. 打开 控制台中的 Plaud,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.mcp-dir%2Fplaud-mcp/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": {
    "plaud": {
      "type": "http",
      "url": "https://api.mcp.ai/p_plaud"
    }
  }
}

README

Plaud

Your Plaud recordings in Claude, ChatGPT and AI agents

Bring your Plaud voice recordings into your AI: list recordings, read speaker-attributed transcripts and the AI-generated summaries and action items. Connect your Plaud account in one click. Read-only, nothing is changed on your Plaud account.

  • 🎙️ Recordings as text: list recordings and read full transcripts
  • 🗣️ Speaker-attributed transcripts: who said what
  • ✅ AI summaries and action items for each recording
  • 🔒 Read-only · 🎁 Free
  • 🔗 One-click connect: authorize your Plaud account, no API keys
  • 💬 Works with any MCP client: Claude Desktop, Cursor, VS Code, Cline, Continue

Versão em português · Full docs (PT-BR) · Agent skill


One-click install

Claude (Web and Desktop)

Anthropic unified MCP installation at claude.ai/customize/connectors. The same link works for Claude Web and Claude Desktop (just be logged in):

➕ Open in Claude and connect

Manual (if the deeplink does not open): claude.ai/customize/connectors → + → Add custom connector → paste Name Plaud and URL https://api.mcp.ai/p_plaud.

Cursor

➕ Install Plaud in Cursor

VS Code (Copilot Chat)

➕ Install Plaud in VS Code

ChatGPT, Manus, OpenClaw and 40+ other clients

Works with any MCP client that speaks MCP over HTTP. The server URL is always:

https://api.mcp.ai/p_plaud

Per-client details: INSTALL.md.


Example prompts

List my latest Plaud recordingsSummarize my last meeting recordingWhat were the action items from yesterday's call?Show the transcript of my recording from this morning

5 tools available

ToolDescription
plaud_list_recordingsLista gravações Plaud. Filtros opcionais: query (substring do nome, case-insensitive), date_from/date_to (YYYY-MM-DD, inclusivo). Com filtro, pagina até 5×100 e retorna todas.
plaud_get_recordingDetalhes de uma gravação Plaud por ID.
plaud_get_notesNotas geradas por IA de uma gravação: resumo compacto, action items e tópicos-chave.
plaud_get_transcriptTranscrição completa com timestamps e atribuição de falantes de uma gravação.
plaud_get_userDados do usuário Plaud autenticado.

Details for each tool: docs/ferramentas.md (PT-BR)


Pricing

Free.


Privacy & data protection

  • Read-only, no tool changes data at the source.
  • Sub-processors: Plaud, the LLM host you choose (Claude, ChatGPT, Cursor, your own agent). Full list in docs/privacidade-lgpd.md.
  • Data returned by the tools is sent to the LLM host you choose, a sub-processor outside our control. We recommend plans with training opt-out.

FAQ

How do I connect? Click connect and authorize your Plaud account. There are no API keys to paste. It is read-only.

Does it change anything in Plaud? No. Every tool is read-only: it lists recordings and reads transcripts, summaries and action items.

How much does it cost? Free.

Is the server open source? The server is proprietary (hosted). This repository is the public wrapper with manifests, docs and skills, all MIT.


Support


License

MIT — see LICENSE. The MCP server at api.mcp.ai/p_plaud is proprietary (hosted); this repo (manifests, docs, skills) is MIT.

来源:README.md,提交 46791cd

工具

11

plaud_list_recordings

Lista gravações Plaud. Filtros opcionais: `query` (substring do nome, case-insensitive), `date_from`/`date_to` (YYYY-MM-DD, inclusivo). Com filtro, pagina até 5×100 e retorna todas.

输入参数pagequeryaccountdate_todate_frompage_size

plaud_get_recording

Detalhes de uma gravação Plaud por ID. Bulk support: accepts file_ids for batched execution.

输入参数accountfile_id必填file_ids

plaud_get_notes

Notas geradas por IA de uma gravação: resumo compacto, action items e tópicos-chave. Bulk support: accepts file_ids for batched execution.

输入参数accountfile_id必填file_ids

plaud_get_transcript

Transcrição completa com timestamps e atribuição de falantes de uma gravação. Bulk support: accepts file_ids for batched execution.

输入参数accountfile_id必填file_ids

plaud_get_user

Dados do usuário Plaud autenticado.

无需输入参数。

show_version

Show Version

Show the current MCP platform and adapter versions.

无需输入参数。

report_bug

Report Bug

Report a bug, missing feature, or send feedback. Include the conversation array with recent messages for reproduction.

输入参数contextmessage必填conversation

connect

Connection Status

Returns connection status and URLs. When all providers are connected, returns authenticated:true and empty pending[]. When credentials are missing, returns connect_url for the toolkit and per-install URLs.

无需输入参数。

toolkit_info

Toolkit Info

Returns the current toolkit state: installed MCPs, their connection status, the accounts connected to each one, and how many catalog tools each exposes.

无需输入参数。

marketplace

Marketplace

The official mcp.ai marketplace — the in-platform catalog of every MCP/tool, AND the way to run them. Covers capability requests like "find an MCP that does X", "consulta um CPF", "is there a tool for Y". Core flow: action=search discovers MCPs by intent → describe returns one MCP's full profile (every tool with its id + params, pricing, auth) so you pick the right tool_id → invoke RUNS that tool. KEY: invoke works even when the MCP is NOT installed — it runs the tool pontualmente (one-off), without adding the MCP to the toolkit and without bloating the tool list. If the MCP needs a credential/login, invoke returns a connect link; if it is paid and the wallet is empty, invoke returns a checkout/top-up link (the user opens it, then you retry). Use install only to make an MCP PERMANENT in the active toolkit (its tools then show up natively in future sessions); prefer invoke for a single/occasional use. list_tools lists what is callable right now. subscribe/cancel handle per-MCP billing; report_bug sends feedback; request_mcp asks us to build a NEW MCP when nothing fits. Search/describe flag installed_in_toolkit vs installed_in_workspace. Writes (install/uninstall/subscribe/cancel and the one-off install behind invoke) require workspace owner/admin. It also carries the mcp.ai PROMPT LIBRARY, which is about ready-made prompt TEXT rather than MCPs: search_prompts finds one, get_prompt returns its full text with {{variables}} filled, and publish_prompt saves a prompt and returns a shareable mcp.ai/p/<slug> link that opens without login.

输入参数limitqueryactionmcp_idmessagetool_idargumentsimmediatetier_slugprompt_bodyprompt_slugprompt_tool另有 11 个

authenticate

Authenticate

MCP.AI for IDE agents (Cursor, etc.): log in in the browser, copy the access token. Best: add it to this server's config as a header `Authorization: Bearer <token>` for a permanent, non-expiring connection. Or paste it here for a session-only login: call with { token: "<jwt>" } after the user pastes, or with no args to get the link.

输入参数token

版本历史

1
  1. v0.1.0最新Sep 16, 2026

评分

已满足适用检查项的 100%

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