Skip to content

Security: kortix-ai/agent-computer-use

Security

SECURITY.md

Security Policy

Reporting a Vulnerability

If you discover a security vulnerability in agent-computer-use, please report it responsibly:

  1. Do not open a public GitHub issue
  2. Email the maintainers directly (see Cargo.toml for contact)
  3. Include steps to reproduce and potential impact

We will respond within 48 hours and issue a fix as soon as possible.

Scope

agent-computer-use requires Accessibility permissions to function. This is an OS-level security boundary — the user must explicitly grant access. agent-computer-use does not:

  • Transmit data over the network
  • Store credentials or sensitive data (refs cache contains only element names/roles)
  • Execute arbitrary code from remote sources
  • Bypass OS security boundaries

Permissions

agent-computer-use uses the following OS permissions:

Permission Why Platform
Accessibility Read/write the accessibility tree, simulate input macOS
AT-SPI2 Read/write the accessibility tree Linux
UI Automation Read/write the accessibility tree, simulate input Windows

There aren't any published security advisories