Intent, verified.

We founded SpecStory to preserve human intent: what people were trying to accomplish with coding agents, why it mattered, and the decisions code alone could not explain. Saving conversations was our starting point.

Our work with agents on the frontier has progressed rapidly from writing better prompts, to supplying better context, to designing the loop through which work gets done. It works so well that we don't question whether agents can produce substantial software anymore.

SpecStory's open source still preserves the thinking behind your work. RunStory independently runs the software against your change and returns evidence to the coding agent. An agent saying "done" is a claim that needs to be verified.

Every change should leave behind something useful for checking the next one. What passed before needs to remain true, unless someone explicitly decides otherwise.

For agentic engineering to succeed, we must connect what someone meant to evidence of what the software actually does.

We started by preserving what people meant. We are building RunStory to hold software to it.