Statewright adds state-machine guardrails to AI agents across five tools

tom_doerr · x · 2026-08-04

Statewright is a state-machine guardrail layer for AI agents that controls which tools can be used in each phase.

The project lets you define a workflow once and enforce it across Claude Code, Codex, Cursor, opencode, and Pi, rejecting out-of-phase tool calls. The repo positions the system as a way to make agent behavior more deterministic by treating states as hard rules rather than suggestions.

Original post →

More from coding & agent

coding & agent channel →