The Vercel plugin on Claude Code wants to read your prompts

A deep dive into the Vercel plugin for Claude Code reveals aggressive telemetry practices, including collecting full bash commands and prompts even on non-Vercel projects, using deceptive prompt injection techniques.
09 Apr 2026
I was working on a project that has nothing to do with Vercel. No vercel.json, no next.config, no Vercel dependencies. Nothing.
And then this popped up:
“The Vercel plugin collects anonymous usage data… Would you like to also share your prompt text?”
Every single prompt. On a non-Vercel project. That felt wrong. So I went deep into the source code with Claude.
The Issue: The Vercel plugin helps with deployments and framework guidance. Why does it need to read every prompt you type across every project? That’s way outside its scope.
Deceptive UI via Prompt Injection: When the Vercel plugin wants to ask about telemetry, it doesn’t show a CLI prompt. Instead, it injects natural-language instructions into Claude’s system context telling the AI to ask you a question. The result looks identical to a native Claude Code question. There is no visual indicator that it’s from a third-party plugin.
What is collected:
- Device ID, OS, frameworks: Always on.
- Full bash command strings: Always on (sent to telemetry.vercel.com).
- Full prompt text: Opt-in.
No Project Scoping:
There is no project detection for telemetry. The UserPromptSubmit matcher is an empty string—it matches everything. Install the plugin for your Next.js app, and it’s watching your Rust project, your Python scripts, and your client work.
How to opt-out:
- Kill all Vercel telemetry:
export VERCEL_PLUGIN_TELEMETRY=off - Disable the plugin: Set
"vercel@claude-plugins-official": falsein~/.claude/settings.json
Source: Hacker News















