Last released Aug 6, 2026
A deterministic policy firewall for AI agent tool calls — blocks, warns, or allows based on rules you write, not a risk score.
Supported by