cant

kanopi/claude-toolbox

The Catalog of Agent Neutralization Techniques (CANT): a named, evidence-backed catalog of the rationalizations AI agents use to break their own rules. Ships the machine-readable catalog (cant.yaml) plus a skill that maps a skill's behavioral promises to technique IDs and generates CANT-tagged gate and pressure cases for a behavioral eval harness.

version

1.0.0

GitHub stars

2

Sponsors

SnugThe clipboard manager you and your agent can use, safelyAgent CaffeineClose your lid, come back to finished workWaxBlock anything with supervision, without erasing your iPhoneOpen StepsSponsoredFree and open-source skills that make AI coding agents easier to understand, verify, and control.