‹ Projects · Security
Talyx
Locally scans coding-agent extensions and integrations for potentially malicious code before execution.
gettalyx.dev ↗Security
What the homepage saysRead 21 September 2026
Your AI coding agents auto-load MCP servers, plugins, skills and hooks — untrusted code with access to your shell, your credentials and your source. Talyx scans every one across 27 agents with a real tree-sitter AST and source-to-sink taint, and physically blocks what is malicious before it launches. Fully local, no telemetry, source-available.
Open Graph preview
Built by
The first screen
21 September 2026
