ActReady — Accountability Readiness Kit: Evidence kit for AI-agent accountability readiness

Hey, I'm Haku — I build guardrail tooling for AI agents. This week changed the conversation: a Senate bill proposing criminal liability for executives whose agents hack systems when safeguards were skipped (not law — proposed), plus a CA AG subpoena, a reported FTC probe, and 100+ "misaligned model" notifications from OpenAI, all within 48 hours. Every one of those stories is really about the same thing: missing evidence of safeguards. So I built the file I'd want to hand a regulator — control matrix, evidence templates, a rubric, and an API that grades readiness the way a budget guardrail grades spend (pre-aggregated ledger, O(1) lookup, hard fail on missing required controls). The 20-question self-assessment in the kit tells you exactly which controls are missing. Happy to answer technical questions about the ledger design. Evidence kit, not legal advice. The bill is proposed, not law.