HanoLab Voice Conversion

io.github.obrr-hhx由 GitHub 帳號 obrr-hhx 發布v1.0.2更新於 Oct 24, 2026

Convert uploaded speech or singing with your saved HanoLab voices, using OAuth and credit quotes.

已驗證Streamable HTTP可網頁執行媒體與設計AI 與機器學習

概覽

AI 產生的概覽

HanoLab 遠端 MCP 伺服器，讓助理用你已儲存的聲音轉換上傳的語音或歌聲，透過 OAuth 登入並依點數報價執行。

可以做什麼
  • 查看已連線帳戶、方案與可用點數
  • 列出已儲存的聲音與最近上傳的音訊，並建立一小時有效的私密瀏覽器上傳連結
  • 為選定音訊與自有聲音準備固定點數報價
  • 在使用者確認預算後提交轉換，輪詢任務狀態並回傳完成的音訊連結
  • 檢視透過此連線提交的最近轉換紀錄
適用情境
適合讓助理用帳戶中已儲存的聲音對錄製的語音或歌聲執行 HanoLab 聲音轉換，並在消耗點數前先提供報價與預算確認。它不會複製新聲音，也不處理購買或訂閱變更。
執行需求
需在支援遠端 MCP 與 OAuth 的用戶端加入遠端 Streamable HTTP 端點，並完成 HanoLab 登入與授權。需要既有的 Starter、Creator 或 Pro 帳戶。上傳透過私密瀏覽器連結進行，工具無法直接讀取本機路徑或聊天附件。
安裝前請注意
轉換會消耗 HanoLab 點數；提交前請確認報價金額與預算上限，中斷後沿用同一報價以免重複扣款。上傳限制為 15 分鐘、100 MB，MP4 輸入回傳的是音訊而非替換影片音軌。僅使用你擁有或獲授權的聲音與音訊，結果會依隱私政策分享給所連線的 AI 主機。

安裝

在 SourceWeft 中

  1. 開啟 儀表板中的 HanoLab Voice Conversion，將其新增到工作區。
  2. 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.obrr-hhx%2Fhanolab/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。

檢視代理程式所依據的安裝指南

其他 MCP 客戶端

把它新增到你客戶端的 mcpServers 設定中。

{
  "mcpServers": {
    "hanolab-voice-conversion": {
      "type": "http",
      "url": "https://hanolab.com/mcp"
    }
  }
}

README

HanoLab integrations

Voice conversion for recorded speech and singing, with private saved voices, credit quotes, and resumable results.

HanoLab · Remote MCP guide · CLI guide · Privacy · Support

This repository distributes public connection metadata, two agent skills, and the client source from the public CLI release. The hosted service implementation is maintained separately.

Remote MCP

Add https://hanolab.com/mcp as a Streamable HTTP MCP server in a client that supports remote MCP and OAuth. Complete the HanoLab sign-in and consent flow. Reading a skill or copying a configuration does not connect your account or install tools.

The remote MCP currently converts uploaded recordings using voices already saved in your account. It can inspect credits and voices, create a private upload link, quote a conversion, submit an approved conversion, and recover its status or output. It does not clone a new voice or expose purchases or subscription changes.

The JSON shape varies by client. mcp.json is an HTTP-style example; use the host's own remote-server setup instructions when necessary.

Example requests after connecting:

  • "Show my saved voices and credits. Do not start a conversion."
  • "Give me a private upload link for my recording, then quote conversion using my saved voice. Wait for my budget confirmation."
  • "Recover the conversion we already submitted and return its result. Do not start a new paid task."

Agent skills

Install both public skills using the Skills CLI:

sh
npx skills add hanolab-ai/hanolab-integrations

Or select one:

sh
npx skills add hanolab-ai/hanolab-integrations --skill hanolab-mcpnpx skills add hanolab-ai/hanolab-integrations --skill hanolab
  • hanolab-mcp: the connected remote MCP workflow, private browser uploads, exact quotes, confirmed credit budgets, and resuming existing tasks.
  • hanolab: local terminal workflow for music generation, voice cloning/conversion, stem separation, mastering, and resumable downloads. Requires a separately installed CLI and local file access.

Local CLI

Requires Node.js 20+ and a HanoLab API key. Install the current release and authenticate privately:

sh
npm install -g https://hanolab.com/downloads/hanolab-cli-0.2.2.tgzhanolab auth loginhanolab help

The CLI client source is copied from the public 0.2.2 release. Release checksums identify the exact downloadable artifact. The client sends authenticated API requests to HanoLab and streams audio uploads/downloads; inspect the source and the privacy policy before connecting your account.

Create the key on the account page and enter it at the hidden terminal prompt. Do not paste credentials into a chat or a project file.

Account access and cost

MCP and API access currently use an existing Starter, Creator, or Pro account. Audio processing consumes the account's credits; checking credits, saved voices, and a quote does not execute a conversion. Follow the server's measured quote and the user's explicitly authorized maximum budget. A successful connection does not purchase a plan or supply unlimited processing.

Uploads through remote MCP support audio or MP4 soundtracks up to 15 minutes and 100 MB; MP4 conversion returns audio rather than a replaced video soundtrack. Only use voices and audio you own or have permission to use. Results are shared with the connected AI host as described in the privacy policy.

Registry metadata

server.json describes the hosted remote endpoint for the official MCP Registry. A source repository or configuration file by itself does not indicate registry approval, third-party listing, or compatibility with every AI host.

Updating

Keep these public skills aligned with HanoLab's documented tools and current CLI release. Publish a new registry metadata version when its public fields change. Do not add secrets, review-account credentials, signed private audio links, or hosted backend source to this repository.

來源：README.md，提交 5282218

工具

9

get_profile

Connected account

Identify the HanoLab account connected to these tools.

無需輸入參數。

get_credits

Credit balance

Read the connected account’s plan and available credits. Does not purchase credits.

無需輸入參數。

list_voices

Saved voices

List the connected user’s own saved voices and whether they are ready. If none are ready, open voicelab to create one.

無需輸入參數。

list_audio

Uploaded audio

List recent audio uploaded through this connection. Use the exact audio_id, never a local path or an arbitrary URL.

無需輸入參數。

prepare_voice_conversion

Estimate voice conversion

Prepare a fixed quote for uploaded audio and a ready owned voice. No credits are spent. Explain estimated_credits, obtain the user’s budget and explicit confirmation, then use convert_voice with this exact quote_id.

輸入參數audio_id必填voice_id必填full_song

convert_voice

Convert voice

Spend HanoLab credits to submit the prepared conversion ONLY after the user confirms its cost and budget. Reuse the exact quote_id after interruption: repeated calls return the original task without another debit. A failed task needs a separate user decision before a new quote.

輸入參數quote_id必填confirmed必填max_credits必填

get_task

Conversion status and results

Check an owned task and return completed audio links. Refreshes worker status and may persist completion or a failure refund. Poll about every 10 seconds; do not submit a new paid task while waiting.

輸入參數task_id必填

list_conversions

Recent conversions

Find recent conversions submitted with this connection. Use get_task to refresh a task. This does not start new tasks.

無需輸入參數。

版本歷史

2
  1. v1.0.2最新Oct 14, 2026
  2. v1.0.0Oct 14, 2026

評分

已滿足適用檢查項的 90%

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
  • 已列出工具必需15 分
  • 描述必需15 分
  • 圖示10 分
  • 文件15 分
  • 授權條款10 分

    請在儲存庫或註冊中心條目中宣告授權條款。