An AI changed your code. RunBoth runs both versions and tells you what actually behaves differently, including the functions nobody touched.

runboth.dev · Source · PyPI

A diff tells you the text changed. It does not tell you the behaviour changed.

RunBoth checks out both versions of your code, generates inputs for every changed function from its signature and from the constants mined out of its own bytecode, runs both versions in separate sandboxed subprocesses, and compares seven observation channels. When they disagree it hands you the exact input that separates them.

No test suite required, because the old version is the specification. No network calls, no AI model, no dependencies.

pip install runboth
runboth install-hook

Three verdicts and no fourth. changed, with the call that proves it. no_change at budget N. abstained, with the reason. It never says safe, because sampling finds differences and cannot prove their absence.

Measured before it was published. 2,548 functions across eight public repositories it had never been tuned on, zero false positives, with an automated oracle built to catch the tool lying. The harness and the raw records ship in the repository so you can check the numbers rather than trust them.

Built by Kyle Clouthier at Clouthier Simulation Labs.