The anti-detect browser your AI assistant drives over MCP
One MCP command connects Claude Code, Cursor or any MCP client to Veilus. The assistant does the work across your profiles; you approve once.
01
You give the orderone sentence 02
The AI writes a scriptsave_script 03
Test run on 3 profilesrun_script repeat until it works
04
Reads errors, fixes itselfget_run_result 05
You approveVeilus Flow 06
Schedule the whole batchcreate_schedule Connect
One MCP command
veilus mcp
Claude CodeCursorany MCP client
your machine only · token required
See the whole loop in 55 seconds
The video is in English. Playing it loads YouTube.
Connect in one command
Turn on the local API in the app, create a token on the API & MCP page, then register Veilus with your client. The app fills in the real path and token for you.
Claude Code
claude mcp add --scope user veilus -e VEILUS_API_TOKEN=<token> -- "<path to Veilus>" mcp
Cursor / Claude Desktop
{
"mcpServers": {
"veilus": {
"command": "<path to Veilus>",
"args": ["mcp"],
"env": { "VEILUS_API_TOKEN": "<token>" }
}
}
} What the assistant can do
- 01Profiles: list, create and launch profiles, each with its own fingerprint
- 02Proxy: import proxy lists, test them, assign pools to profiles
- 03Page driving: navigate, click, type and read a running profile's page
- 04Scripts: save, compile and run Playwright scripts, read per-profile results
- 05Batch and schedules: run on many profiles at once or every day at a set time, after you approve the script
- 06Datasets: give each profile its own rows of accounts, posts or keywords
Scripts added over MCP run on at most 3 profiles until you approve them in the app. The API listens on your machine only.
next step
Download Veilus. Connect your AI assistant. Give the order.
Download & try Pro free for 7 days No card needed · profiles stay on your machine by default