Statewright – Visual state machines that make AI agents reliable
Forces 13B models to solve SWE-bench tasks by making the problem smaller, not the model bigger.

Protocol-level tool enforcement prevents jailbreaks better than system prompts.
Engineers building reliable AI agent workflows
LangGraph · Microsoft AutoGen · Temporal
Forces 13B models to solve SWE-bench tasks by making the problem smaller, not the model bigger.
Two-commit spec repo with no working implementation or demo.
TDD state machine leash for Claude Code avoids agent drift, but niche audience.
465 passing tests, zero deps, guards extracted from 5,000+ real agent executions.
Formal verification for LLM workflows—CTL model checking, Z3 proofs, zero hallucination math.
Compile-time generated scheduler beats manual match-loop-state hell for complex state machines.