Why it exists
Agent work is hard to trust when each Run stands alone. A transcript can show what happened, but it often leaves ownership, context, cost, Review, and the next lesson scattered across different tools. Rudder gives people and Agents a shared operating record. An organization has a goal, durable Agents, Chat and Issue work surfaces, projects, runtime access, budgets, and review paths. Output and feedback stay connected to the work that produced them.What Rudder does
Rudder coordinates work around Agent runtimes. It records the task, wakes the selected runtime, preserves execution evidence, and keeps Review, spend, and follow-up visible. Chat works well for conversation-led tasks. Issues add explicit status, ownership, dependencies, and Review when that structure helps. Both can carry real work from a goal to an inspectable result.What Rudder does not do
Rudder is not an Agent runtime, a generic social chat product, or a complete replacement for delivery tools such as GitHub or Jira. It connects those parts of the work loop and leaves execution to the configured runtime.The work loop
Quick Start
Install Rudder and open the local app.
Core concepts
Learn how organizations, Agents, work surfaces, Runs, and Reviews fit together.