Claude Workbench banner
Norman-else Norman-else

Claude Workbench

AI community

Description

A modern GUI management tool for the Claude Code CLI with a built-in AI Assistant. It provides a visual interface to manage MCP servers, API credential profiles, custom slash-commands, agent skills, and community plugins — all without manually editing JSON or shell config files. The integrated AI Assistant lets you interact with Claude directly.

Installation

This entry records only its repository, not the path inside it, so there is no exact command to give. Open the source below and copy the folder into ~/.claude/skills/, or the file into ~/.claude/agents/.

README

Claude Workbench

A modern GUI management tool for the [Claude Code CLI](https://docs.anthropic.com/en/docs/claude-code) with a built-in **AI Assistant**. It provides a visual interface to manage MCP servers, API credential profiles, custom slash-commands, agent skills, and community plugins — all without manually editing JSON or shell config files. The integrated AI Assistant lets you interact with Claude directly inside the app, with support for tool use, file/image uploads, web search, and streaming responses.

Available as a **web app** (runs in your browser) or a **desktop app** (native Electron wrapper for Windows and macOS).


Features

AI Assistant

  • Built-in conversational AI assistant powered by Claude, accessible from any tab via floating panel
  • Real-time streaming responses with smooth token-by-token rendering
  • Tool use: Slash commands (/tool-name) to invoke Claude Workbench management tools directly from chat — browse MCP servers, manage environment profiles, explore skills and plugins
  • File & image uploads: Attach images (PNG, JPEG, GIF, WebP), PDFs, and text-based files (Markdown, JSON, YAML, code files) via paperclip button or paste from clipboard
  • Web search: Claude can search the web in real time when answering questions about current events or external topics
  • Multi-conversation management: Create, switch, rename, and delete conversations with auto-generated titles
  • Model selection: Switch between available Claude models on the fly, including via /model-name slash command
  • Stop generation: Abort in-progress responses at any time with the stop button; partial output is preserved
  • Auto-continue: Long responses that exceed token limits are automatically continued without user intervention
  • Light & dark theme: Full theme support matching the rest of the application

Environment Profiles

  • Create and manage multiple named API credential profiles (e.g., Production, Development)
  • Configure `A