imprint
by kessler-frost
Overview
Enables AI agents to programmatically control a real terminal, interact with TUI applications, and capture pixel-perfect screenshots for visual analysis and testing.
Installation
imprint --port 8080Environment Variables
- IMPRINT_INSTALL_DIR
- SHELL
Security Notes
The server's core functionality involves exposing a pseudo-terminal (PTY) where AI agents can execute arbitrary shell commands via 'ttyd'. This is an inherent risk by design; the security of the host machine relies on the agent's trustworthiness and the sandbox it operates within. The 'go-rod' library uses 'page.Eval' to interact with the 'xterm.js' terminal emulator in the headless browser, which is confined to the browser's JavaScript environment and not executing arbitrary code on the host OS. The REST API and 'ttyd' bind to '127.0.0.1' (localhost) by default, limiting direct external network exposure. The CORS middleware allows 'Access-Control-Allow-Origin: *', which would be a concern if the server was publicly exposed, but is less critical for a local-only daemon. No obvious hardcoded credentials or unintended critical vulnerabilities were found, beyond the explicit design choice of providing a programmable shell to an AI agent.
Similar Servers
UI-TARS-desktop
A desktop GUI automation and AI agent application that allows users to control their computer using natural language, integrating various multimodal AI models and supporting local and remote operation.
chrome-devtools-mcp
Allows AI coding agents to control, inspect, and debug a live Chrome browser for automation, performance analysis, and in-depth debugging.
Windows-MCP
Enables AI agents (LLMs) to interact with and automate tasks on the Windows operating system, including UI interaction, file navigation, application control, and QA testing.
mcp-server-browserbase
Enables LLMs to perform cloud browser automation tasks such as navigating, interacting with elements, extracting data, and capturing screenshots on web pages.