Paper Lantern
About
Paper Lantern setup CLI - one-line install to configure Claude Code, Cursor, Windsurf, Copilot, Codex, and Gemini CLI with Paper Lantern
Details
- Author
- paperlantern-ai
- Downloads
- 381
- Categories
- Developer Tools
Jump to
- One‑command setup via npx paperlantern@latest
- OAuth authentication for secure access
- Automatic MCP server configuration for supported agents
- Agent rules installed so agents know when to use Paper Lantern
- Five tools for research exploration, deep dives, comparisons, feasibility checks, and feedback
Setting up with Highlight
This MCP is not yet compatible with Highlight’s one-click setup. However, you can still use it with Highlight by following these steps:
- Download and install Highlight from highlightai.com/download
- Navigate to the plugins tab and select "Add Custom Plugin"
-
Configure the plugin with the settings below
Plugin Name
Paper LanternCommand (node, npx, python, etc.)Please refer to the README for specific instructions on how to obtain API keys or other required environment variables.
- Enable "Start Automatically" if you want the plugin to start when Highlight launches
From the repository
Run npx paperlantern@latest in your terminal, then select your preferred AI agents and log in through OAuth. The setup CLI automatically configures the MCP server and installs agent rules.
Claude Desktop / Cursor
Paste into your MCP client config file to install this server.
{
"mcpServers": {
"paper lantern": {
"paper-lantern": {
"type": "http",
"url": "https://mcp.paperlantern.ai/chat/mcp?key=YOUR_KEY"
}
}
}
}
McpServers
{
"paper-lantern": {
"type": "http",
"url": "https://mcp.paperlantern.ai/chat/mcp?key=YOUR_KEY"
}
}
When your agent uses Paper Lantern
Paper Lantern activates when your agent is making technical decisions where research evidence could improve the outcome - choosing between algorithms, architectures, or techniques.
It does not activate for syntax questions, library lookups, debugging, or general programming tasks.
Sign in to leave a review
Use Google, GitHub, or an email account so ratings stay tied to real people.
No reviews posted yet.





