How Windrunner works

Windrunner separates durable work from the context around it:

Building block	Purpose
Project	Workspace for related work.
Work item	A task, question, approval, review, or decision.
Entry	An update, answer, evidence, proposal, or resolution attached to a work item.
Relationship	A typed connection such as a blocker or dependency.
User and team	People who own, contribute to, or access work.
Conversation	A user-owned AI Agent session with optional workspace context.

A simple workflow

1. Create a work item for the outcome or question.
2. Add entries as the work develops.
3. Add relationships when a dependency, blocker, or supporting connection is important.

For example, use a QUESTION work item, add EVIDENCE and ANSWER entries, then mark the answer as accepted.

AI Agent can read these records, summarize them, and suggest changes. Proposals remain available for review until a person accepts or rejects them.

How it works

Everything connects through the Work Graph.

People and teams get access to projects, projects organize work, and every update, dependency, and decision stays connected for humans and the AI Agent.

Record relationshipAgent, context, or activity flow

The Work Graph keeps access, hierarchy, context, ownership, and time connected while every AI change remains reviewable.

See the focused pages for work items, entries, relationships, and projects.