Kling

io.github.mindstone由 GitHub 帳號 mindstone 發布v0.5.0更新於 Nov 3, 2026

用於模型上下文協議主機的 Kling AI 影片產生 MCP 伺服器

AI 翻譯由 AI 模型根據下方的原文描述翻譯,可能有誤。

原文描述

Kling AI video generation MCP server for Model Context Protocol hosts

已驗證STDIO僅桌面媒體與設計AI 與機器學習

概覽

Kling 的本機 MCP 服務;無需登入;STDIO。

暫無 AI 概覽:以下資訊來自註冊中心條目和服務本身。

安裝

在 SourceWeft 中

  1. 開啟 儀表板中的 Kling,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。

Desktop only,透過 STDIO。 STDIO 服務會啟動本機處理程序,因此需要 SourceWeft 桌面主機。

其他 MCP 客戶端

參照 儲存庫 中的啟動說明。

README

@mindstone/mcp-server-kling

[npm version] [License: FSL-1.1-MIT]

Kling AI video and image generation MCP server for Model Context Protocol hosts. Generate AI videos from text or images, extend them, add lip-sync, generate images, and manage generation tasks through a standardised MCP interface.

Status

Requirements

  • Node.js 20+
  • npm

One-click install

[Add to Cursor] [Add to VS Code] [Add to VS Code Insiders]

After clicking the button, your host will prompt you to fill: KLING_ACCESS_KEY, KLING_SECRET_KEY, KLING_REQUEST_TIMEOUT_MS.

Manual config for Claude Desktop / Claude Code / Goose / Continue.dev (Kling)
json
{  "mcpServers": {    "Kling": {      "command": "npx",      "args": [        "-y",        "@mindstone/mcp-server-kling"      ],      "env": {        "KLING_ACCESS_KEY": "",        "KLING_SECRET_KEY": "",        "KLING_REQUEST_TIMEOUT_MS": "60000"      }    }  }}

Quick Start

Install & build

bash
cd <path-to-repo>/connectors/klingnpm installnpm run build

npx (once published)

bash
npx -y @mindstone/mcp-server-kling

Local

bash
node dist/index.js

Configuration

Environment variables

  • KLING_ACCESS_KEY — Kling API access key
  • KLING_SECRET_KEY — Kling API secret key
  • MCP_HOST_BRIDGE_STATE — optional path to a host bridge state file used for credential management
  • MINDSTONE_REBEL_BRIDGE_STATE — backwards-compatible alias for MCP_HOST_BRIDGE_STATE
  • MCP_WORKSPACE_PATH — optional directory that local media inputs (image_path, audio_path) must live inside. Defaults to the system temp directory when unset. Files elsewhere are refused before any read.
  • KLING_DOWNLOAD_ROOT — optional directory that download_kling_video output paths must live inside. It must itself resolve inside the workspace (MCP_WORKSPACE_PATH, or the system temp directory when unset); anything else is refused with guidance. Default: <workspace>/kling-downloads (auto-created). Sensitive paths (~/.ssh, ~/.aws, /etc, shell rc files) are refused even when the root would otherwise permit them.
  • KLING_REQUEST_TIMEOUT_MS — optional override (positive integer ms, max 30 min) for the outbound HTTP request timeout applied to both Kling API and host-bridge calls. Default: 60000 (60s). Raise this if you see TIMEOUT errors on slow submits; lower it if you want tighter bounds.

Host configuration examples

Claude Desktop / Cursor

json
{  "mcpServers": {    "Kling": {      "command": "npx",      "args": ["-y", "@mindstone/mcp-server-kling"],      "env": {        "KLING_ACCESS_KEY": "your-access-key",        "KLING_SECRET_KEY": "your-secret-key"      }    }  }}

Local development (no npm publish needed)

json
{  "mcpServers": {    "Kling": {      "command": "node",      "args": ["<path-to-repo>/connectors/kling/dist/index.js"],      "env": {        "KLING_ACCESS_KEY": "your-access-key",        "KLING_SECRET_KEY": "your-secret-key"      }    }  }}

Tools (10)

Configuration

  • configure_kling_api_keys — Save Kling API credentials

Video generation

  • generate_kling_video — Create an AI-generated video from a text description
  • generate_kling_image_to_video — Animate a still image into a video (public HTTPS URL or local file via image_path)
  • extend_kling_video — Continue a generated video, adding ~4-5 seconds per call
  • generate_kling_lip_sync — Make a person in a video speak, from text (text-to-speech) or an audio file

Image generation

  • generate_kling_image — Create an AI-generated image from a text description, with optional reference image

Models: the video tools default to kling-v2-6 and the image tool to kling-v2-1. kling-v3 (Kling 3.0) is available on all three generation tools through the model parameter; it costs about twice as many Kling units per second of video and per image, so it is not the default.

Task management

  • check_kling_task — Check if a generation task is complete (task types: text2video, image2video, video-extend, lip-sync, image)
  • list_kling_tasks — List your generation tasks with pagination
  • download_kling_video — Save a generated video or image to a local file (result URLs expire 30 days after generation). Only Kling result URLs (klingai.com hosts) are accepted.

Account

  • get_kling_balance — List resource packages on the account and their remaining quantities

All generation tools accept an optional callback_url (HTTPS) that Kling POSTs to on task status changes, as an alternative to polling.

Licence

FSL-1.1-MIT — Functional Source License, Version 1.1, with MIT future licence. The software converts to MIT licence on 2030-04-08.

來源:connectors/kling/README.md,提交 9b9aa57

工具

0
SourceWeft 尚未在沙箱中啟動過該軟體包,因此其工具未知。

版本歷史

2
  1. v0.5.0最新Nov 3, 2026
  2. v0.4.1Sep 16, 2026

評分

已滿足適用檢查項的 40%

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25 分

    請保持端點能夠回應 MCP 握手;SourceWeft 每 30 天探測一次。

  • 已宣告認證方式10 分

    請在註冊中心條目中宣告憑證請求標頭,或保持端點能夠回應探測,以便判定其認證方式。

  • 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

    請在儲存庫或註冊中心條目中宣告授權條款。