Clipboard Operator

An Electron-based clipboard manager that applies JavaScript transformation operations to clipboard content. Lightweight, customizable, and designed for developers who frequently need to reformat or transform copied text — including AI-powered transformations via LLM APIs.

---

Kudos to @minosvasilias for pointing me in the right direction

---

Video

...loading

---

Features

• Call up an input menu with a keyboard shortcut or tray icon.
• Apply custom transformation operations to clipboard content.
• Define operators via a JSON config file pointing to JavaScript scripts.
• Supports LLM-powered operators using any OpenRouter-compatible model.
• Streams LLM responses token-by-token with a live fading log panel in the UI.
• Ships with an examples/ folder to get started quickly.

---

How to Use

1. Open the input menu with Ctrl + Shift + Space or by clicking the tray icon.
2. Select an operation and press Enter.
3. The clipboard content is instantly updated and the menu closes.

---

Getting Started

On first run, the app creates a workspace/ folder next to the executable (or in the project root when running from source). This folder is your personal configuration space.

Folder structure

workspace/
  operators.json    ← defines your operators
  config.json       ← optional: API keys and model settings

The examples/ folder in the repo contains ready-to-use operators and scripts you can copy into your workspace.

---

Defining Operators

Operators are listed in workspace/operators.json. Each entry points to a JavaScript script:

[
  {
    "operator": "To Pascal Case",
    "description": "Converts clipboard text to PascalCase",
    "icon": "text_fields",
    "aliases": ["pascal", "PascalCase"],
    "script": "./examples/to-pascal-case.js"
  },
  {
    "operator": "Fix Grammar",
    "description": "Fixes grammar and spelling using AI",
    "icon": "spellcheck",
    "aliases": ["grammar", "fix", "spelling"],
    "script": "./examples/openrouter-api-call.js",
    "instruction": "Fix any grammar and spelling mistakes. Keep the same tone and style."
  },
  {
    "operator": "Proofread",
    "description": "Proofreads text and suggests corrections using AI",
    "icon": "rate_review",
    "aliases": ["proofread", "review", "check"],
    "script": "./examples/openrouter-api-call.js",
    "instruction": "Proofread the text. Fix spelling, grammar, punctuation, and awkward phrasing. Keep the original meaning and tone intact. Return only the corrected text."
  }
]

---

Writing a Script

Each script exports an async run function that receives the clipboard content and an operator context object, and returns the transformed result:

export async function run(input, operator) {
    return input.trim().toUpperCase();
}

The operator argument contains the fields defined in operators.json plus two extras injected at runtime:

Field	Description
operator.instruction	The instruction string from operators.json, if set
operator._workspaceDir	Absolute path to the workspace/ folder
operator.onToken	Callback (token: string) => void — call this to stream output tokens live to the UI

Example: To Pascal Case

examples/to-pascal-case.js

export async function run(input) {
    return input
        .trim()
        .split(/[\s_\-]+/)
        .filter(word => word.length > 0)
        .map(word => word.charAt(0).toUpperCase() + word.slice(1).toLowerCase())
        .join('');
}

Input: "hello world" → Output: "HelloWorld" Input: "some_variable_name" → Output: "SomeVariableName"

---

LLM-Powered Operators

For AI-powered transformations, use the shared examples/openrouter-api-call.js script. It calls the OpenRouter API with any model you configure, streaming the response token-by-token. As the model generates text, each token is forwarded live to a fading log panel in the UI so you can see progress in real time.

1. Add your API key

Create workspace/config.json:

{
  "OPENROUTER_API_KEY": "sk-or-...",
  "model": "x-ai/grok-4-fast"
}

You can also set OPENROUTER_API_KEY as an environment variable instead.

2. Add operators with an instruction field

The instruction field in operators.json tells the LLM what to do with the clipboard text:

{
  "operator": "Fix Grammar",
  "description": "Fixes grammar and spelling using AI",
  "icon": "spellcheck",
  "aliases": ["grammar", "fix"],
  "script": "./examples/openrouter-api-call.js",
  "instruction": "Fix any grammar and spelling mistakes. Keep the same tone and style."
}

Multiple operators can reuse the same script — just change the instruction:

{
  "operator": "Proofread",
  "description": "Proofreads text and suggests corrections using AI",
  "icon": "rate_review",
  "aliases": ["proofread", "review"],
  "script": "./examples/openrouter-api-call.js",
  "instruction": "Proofread the text. Fix spelling, grammar, punctuation, and awkward phrasing. Keep the original meaning and tone intact. Return only the corrected text."
}

---

Installation (from source)

git clone <repo-url>
cd clipboard-operator

npm install
npm start

Requirements:

• Node.js (>= 18)
• Electron

Development

# Start with hot reload
npm run dev

# Build for distribution
npm run build

---

FAQ

Q: Can I define my own operators? A: Yes. Add entries to workspace/operators.json and create a script that exports a run function. You can also reuse the LLM script with a custom instruction.

Q: Does it work on Linux/Mac? A: Electron is cross-platform. Tested primarily on Windows.

Q: Which LLM models are supported? A: Any model available on OpenRouter. Set the model name in workspace/config.json.

---

Contributing

Contributions are welcome!

1. Fork the repository
2. Create your feature branch (git checkout -b feature/my-feature)
3. Commit your changes
4. Open a Pull Request

---

License

MIT License — feel free to use, modify, and distribute.