
Podwise
io.github.hardhackerlabs由 GitHub 帳號 hardhackerlabs 發布v1.0.0更新於 Oct 27, 2026
Podcast knowledge for AI agents: search episodes, pull transcripts and ask questions.
概覽
遠端 Podwise MCP 伺服器,讓助理搜尋播客、取得逐字稿與摘要、進行問答,並匯出單集筆記。
- 可以做什麼
- 搜尋播客與單集、查看熱門單集與帳號資訊
- 管理媒體庫:已追蹤的節目與單集、節目單集清單、追蹤與已讀狀態
- 處理或匯入單集,包括上傳音訊
- 讀取單集摘要、逐字稿、章節、重點與心智圖,跨逐字稿提問並翻譯單集
- 建立、列出、匯出與傳送片段,並將單集匯出為 Markdown 或 SRT
- 企業版工具用於音訊處理、翻譯、匯出與用量查詢
- 適用情境
- 適合讓助理追蹤已訂閱節目、摘要或跨單集研究某個主題、產生每週收聽回顧,或把逐字稿整理成筆記、片段與語言學習卡片。
- 執行需求
- 需要 Podwise 帳號,並透過瀏覽器完成 OAuth 2.0 登入,儲存庫中不保存 API 金鑰。需要能連線到遠端端點。部分功能需要 Podwise Pro 或 Enterprise 方案,enterprise 系列工具供企業版 API 使用。當用戶端與瀏覽器不在同一台機器時,OAuth 回呼可能失敗。
安裝
在 SourceWeft 中
- 開啟 儀表板中的 Podwise,將其新增到工作區。
- 連接 所需的 憑證。它們會被私密儲存,絕不會在公開頁面上展示。
- 為需要使用其工具的對話啟用該服務。
Web executable,透過 Streamable HTTP。 遠端服務在工作區中設定後即可從網頁執行環境執行。
交給代理程式安裝
把這段提示詞貼上到 Claude Code、Codex、Cursor 等代理程式,或 SourceWeft 對話中。代理程式會讀取該伺服器的安裝指南,先向你展示發布者、端點和登入方式,經你同意後再把它新增到你的客戶端。金鑰不會寫進任何命令。
閱讀 https://sourceweft.com/mcp/io.github.hardhackerlabs%2Fpodwise/install.md,並按其中說明把這個 MCP 伺服器接入我的客戶端。修改任何設定之前,先告訴我它的發布者、端點和登入方式,等我確認。不要把 API 金鑰寫進命令或我會提交的檔案;問我金鑰應從哪裡取得。其他 MCP 客戶端
把它新增到你客戶端的 mcpServers 設定中。 客戶端會在首次連接時完成 OAuth 登入。
{
"mcpServers": {
"podwise": {
"type": "http",
"url": "https://mcp.podwise.ai/mcp"
}
}
}README
Podwise Plugin
[Claude Code] [Cursor] [Codex] [License: MIT] [AllMCPs]
Turn podcast episodes into AI-powered insights from inside Claude Code, Cursor, Codex and ChatGPT. This plugin bundles the Podwise Agent Skill with the remote Podwise MCP server (https://mcp.podwise.ai/mcp), so installing the plugin wires everything up.
In short: a podcast MCP server for AI agents. Point any MCP client at https://mcp.podwise.ai/mcp and it can search podcasts and episodes, pull a full transcript, read the AI summary, chapters, highlights, Q&A and mind map, ask questions across transcripts, translate an episode, clip key moments, and export notes as Markdown or SRT. The skill teaches your agent how to combine those tools into repeatable workflows.
What's included
Workflows
The skill routes your intent automatically:
MCP server and tools
Endpoint: https://mcp.podwise.ai/mcp — Streamable HTTP, OAuth 2.0. No API key lives in this repository.
Listed on the official MCP Registry as io.github.hardhackerlabs/podwise (see server.json).
38 tools, grouped:
Full parameter reference: skills/podwise/references/mcp.md.
Install
Claude Code
Then authenticate the MCP server when prompted (OAuth runs in your browser via /mcp).
Cursor
Open Customize → Plugins → Add → From GitHub Repository and import this repository as a marketplace (use the full URL):
Then click Add on podwise and authenticate the MCP server when prompted (Customize → Plugins → podwise → Authenticate).
Once Podwise is listed in the Cursor Marketplace, you can also install it from any agent chat:
The MCP server alone is also available as a one-click install from cursor.directory.
Codex / ChatGPT
Then complete the OAuth flow when prompted.
Other agents (skills only)
If your agent reads skills directly but does not install plugins, install the skill and add the MCP server manually:
Then add the server to your MCP configuration. Use the transport name your client expects — http for Claude Code-style clients (as in .mcp.json), streamable-http for Agent Plugins clients (as in mcp.json):
Usage
Once the plugin is installed and authenticated, just ask. For example:
- "Catch me up on this week's podcast episodes"
- "Summarize this podcast episode: "
- "Research what podcasts say about "
- "Write my weekly listening recap"
- "Make language-learning cards from this episode's transcript"
The skill checks the connection with get_me first, then routes the request to the matching workflow.
Requirements
- A Podwise account. The MCP server uses OAuth 2.0; no API key is stored in this repository.
- Network access to
https://mcp.podwise.ai/mcp. - Some capabilities require a Podwise Pro or Enterprise plan; the
enterprise_*tools are for Enterprise-plan API usage. See podwise.ai for plans.
Remote / SSH note: the host's HTTP MCP OAuth uses a loopback callback, which does not work when the host runs on a machine different from your browser. Use a connection method supported by your host, or see the troubleshooting docs at https://docs.podwise.ai. Do not run third-party proxy commands to work around authentication.
Repository layout
The MCP definition is duplicated as mcp.json (portable Agent Plugins, streamable-http) and .mcp.json (referenced by the Claude Code and Codex manifests, http). Keep the transport type matching each platform.
Brand assets
Brand color is #7948E8.
Privacy and support
- Privacy policy: https://podwise.ai/privacy
- Terms of service: https://podwise.ai/terms
- Support and docs: https://docs.podwise.ai
License
MIT © Podwise
來源:README.md,提交 9aa9811
工具
0版本歷史
1- v1.0.0最新Oct 18, 2026
評分
已滿足適用檢查項的 85%
該評分衡量此收錄條目的完整程度,依據註冊中心條目、端點與公開來源計算得出。它不是對伺服器本身的評價,也不表示伺服器安全或目前可用:可連通性是帶日期的檢查結果,評分也不包含安裝量或使用者評價。
- 近 30 天內可連通必需25 分
- 已宣告認證方式必需10 分
- 已列出工具15 分
請讓伺服器在無憑證時回應 tools/list,或發布可在沙箱中無憑證啟動的軟體包。
- 描述必需15 分
- 圖示10 分
- 文件15 分
- 授權條款10 分