First project walkthrough

This walkthrough takes one project from an empty repository to a planned, agent-ready Work Order. It is the end-to-end loop Chiron is built around: knowledge in, plan out, execution shared.

Follow it once and the rest of the docs will make sense as elaborations of these five steps.

1. Set up the project

Create the project and connect its repository as in Getting started. For a brand-new project, an empty repo is fine — Chiron will still model the documents and work items you connect.

2. Let the ontology fill in

Wait for setup_status: ready. Behind that flag, ingestion has parsed the code into modules and symbols, built the graph, generated documentation, and started syncing work items and commits from your provider. Check what landed:

curl "$ONTOLOGY_URL/projects/$PROJECT_ID/knowledge/discover" \
  -H "X-API-Key: $INTERNAL_SERVICE_KEY"

The response lists every source, whether it is indexed or live, and a per-source sync state so you can tell "empty" from "the credential expired." See Discover.

3. Brainstorm a Work Order

Open the project's Plan chat and describe what you want to build in plain language. Chiron proposes a Work Order — carrying the problem, the intent, the requirements, and acceptance criteria — grounded in what it retrieved from the ontology. When the brief is ambiguous, it proposes the answer it would pick, marks it a suggestion, and stops for a human.

4. Send it to execution

Approve the Work Order and send it to the board. Its status moves ready -> in_development and it lands in the Backlog column of the execution board, where people and agents pick up work the same way.

5. Assign an agent

Assign the Work Order to a person or their agent. A daemon agent running through the Chiron CLI reads the Work Order, pulls the project's memory and ontology context, and starts — spending fewer tokens because it is not rebuilding context it already had.

That is the loop. Everything else in these docs deepens one of these steps.