Quick start
Create an account, join an organization, and open your first project.
This guide assumes a local development stack. For production URLs, use the values your administrator provides.
Prerequisites
- Ragnarock web app running (default
http://localhost:3000) - Backend API running (default
http://localhost:8000) - Docker services: Postgres, Redis, Typesense (see Local setup)
1. Configure the frontend
Create ragnarock/.env.local:
NEXT_PUBLIC_API_URL=http://localhost:8000The API client sends session cookies to this host. Better Auth lives at http://localhost:8000/api/auth (not under /api for all REST routes).
2. Sign in
- Open
http://localhost:3000. - Go to Sign up (
/sign-up) or Sign in (/sign-in). - Complete email verification if your deployment requires it.
- If prompted, select or create an organization.
Details: Create account.
3. Create or open a project
- Navigate to Projects —
/dashboard/projects. - Create a project or open one you were invited to.
- Open Members and confirm your persona matches how you will use AI (developer vs business owner, etc.).
4. Run through the core workflow
Ragnarock AI — /dashboard/projects/{projectId}/ragnarock
Start a chat session. Ask for a requirements outline or implementation question.
Documentation — .../documentation
Create an SRS or PRD. Use AI-assisted edits where available.
Repositories — .../repositories
Link GitHub (after org integration) so developer intelligence can use code context.
Tasks — .../tasks
Track work items, or run the planner agent from the API to generate a backlog.
Linear — .../linear
Connect and sync issues if your team uses Linear.
5. Invite your team
- Members → invite by email.
- Assign personas per role.
- Share this documentation site for shared vocabulary.
API quick reference
| Action | Method | Path |
|---|---|---|
| List projects | GET | /projects |
| Ragnarock chat | POST | /projects/:projectId/ai/ragnarock/chat |
| List documentations | GET | /projects/:projectId/documentations |
Full reference: Backend and Scalar at http://localhost:8000/docs.
AI features need Redis, the agent worker (:8100), and valid LLM keys in ragnarock-agent/.env.
If chat returns 202 but never completes, see Common issues.