Who it is forA developer or analyst who asks an agent to handle PDFs but does not want the files sent to a web service.
The problemMost agent PDF workflows rely on online converters or ad hoc Python scripts that break on odd files.
The outcomeAn agent that can merge, split, OCR and compress PDFs through a local stdio server.
What you need first
- The paperpilot or paperpilot-mcp binary from GitHub Releases, or Rust 1.75+ to build from source
- An MCP client such as Claude Code, Codex, Cursor or Claude Desktop for agent use
- Node.js and pnpm only for the desktop app
Steps
- Download paperpilot-mcp from the GitHub Releases page
- Print a ready config for your agent with
paperpilot-mcp --client-config codex(or claude-code, cursor, claude-desktop) - Merge the printed config into your agent's MCP settings and restart it
- Try: Merge these two PDFs and compress the result
Prompt for your agent
Install KrushnaVardhanReddy/PaperPilot's paperpilot-mcp, add it to <my agent> with the config from --client-config, then merge <a.pdf> and <b.pdf> and compress the result. Keep all files local.
Caveat. This card is written from the KrushnaVardhanReddy/PaperPilot README, not from a test run by Deduped. Before you start, make sure you have the paperpilot or paperpilot-mcp binary from GitHub Releases, or Rust 1.75+ to build from source. Check the repo for newer instructions before you rely on these steps.
Use this in your AI agent
Ask your agent
Using the Deduped MCP, get the setup steps for KrushnaVardhanReddy/PaperPilot and "Give Claude Code or Codex local PDF tools over MCP"
Add Deduped to Claude Code once
claude mcp add --transport http deduped https://dedupedai.com/mcp
Every Monday: the 5 hottest repos and exactly how to use them. Free, one email a week, unsubscribe anytime.
Get them freeShare: X LinkedIn Hacker News Email Copy link
Third-party repos we don't own. AI-written use cases can be wrong. Disclaimer