Deterministic pre-flight policy check for AI agent actions, run by a Prolog rule...
https://api.x402ai.dev/api/policy-checkOverview
Deterministic pre-flight policy check for AI agent actions, run by a Prolog rule engine (SWI-Prolog), no LLM. POST JSON {profile: baseline or strict, context: {workdir, home, sandboxed}, actions: 1-20 of read, write, delete, exec, network, pay, self_modify}; returns {policy, results, summary, checked_rules}: one verdict per action (allowed, refused, needs_review) with the rules that decided it. Same input, same output. Checks the actions it is sent; does not enforce them. Typically under 1 s.