> ## Documentation Index > Fetch the complete documentation index at: https://docs.allgoodhq.com/llms.txt > Use this file to discover all available pages before exploring further. # Orchestration > How allGood decides what happens to a person: policies produce decisions, decisions produce actions, and every step is recorded. Orchestration is the part that decides. A **policy** reads the data layer and writes a decision. Policies chain into a graph, so one policy's output is the next one's input. A **decision** is the record of what was chosen for one person and why. An **action** is the decision carried out — an email sent, a field written, a record pushed to your CRM. The payoff of writing it down this way is that every outcome is explainable after the fact. "Why did this person get this email?" has an answer you can read. Start with [Policies, plans and motions](/mk/orchestration/policies-plans-and-motions) for how the layers fit together. ## Pages in this section Being written. Each page below will land as its own file in this tree. * **Policies** — what a policy is, its declared inputs and outputs, and how gates work * **Policy graph** — how policies chain, and how to read the dataflow map * **Decisions** — the decision log: what was decided, for whom, and why * **Actions** — how a decision becomes something that actually happens * **Plans** — the state machine underneath a running motion ## Related What policies read from — people, companies, activity and records. The campaign layer that sits on top of plans and policies. Questions? Reach out to [support@allgoodhq.com](mailto:support@allgoodhq.com).