First login
The first time you open a fresh Exolvra instance, it walks you through a short setup. This page explains what each screen asks for and why.
Create your admin account
On a fresh instance, opening the dashboard takes you to the Welcome screen. There’s no preset password to hunt for in logs — you create the administrator account right there: enter your email, pick a password (at least 8 characters), and click Create account & continue. That’s the one and only time this screen appears; the moment the account exists, the welcome route closes permanently and all future visits go to the normal sign-in page.
Open http://localhost:5079/ and follow the prompt.
The setup wizard
After creating your account you land on the onboarding wizard — five short steps that take about two minutes. You can skip it at any point with the Skip for now link at the bottom; every choice is editable later from Settings.
Step 1 — About you
Enter your display name, your role, and how you want agents to communicate with you (concise, detailed, casual). Agents use this to tune their tone. It’s stored on your profile and can be changed later from Profile → About you.
Step 2 — Trial & license
Every fresh install runs the full product on a free 7-day trial — nothing is locked, so you can skip straight past this step and decide later. If you already have a license key, paste it here. If you want more time to evaluate, enter your email and click Get 14-day key — a free 14-day trial key is issued and applied automatically. See Licensing for how the trial works.
Step 3 — Pick a model provider
Two ways to connect a model — pick whichever you already have:
An API key. Choose a provider and paste a key:
| Provider | Where to get a key |
|---|---|
| Anthropic | console.anthropic.com |
| OpenAI | platform.openai.com/api-keys |
| Google Gemini | aistudio.google.com/app/apikey |
| Groq | console.groq.com/keys |
| Mistral | console.mistral.ai |
| Ollama (local) | No key — just point at your local Ollama URL |
A subscription CLI. If you have a Claude, ChatGPT, or Google subscription, you can use it instead of an API key: the wizard detects installed coding CLIs (and can install them), signs you in, and sets the result as your default model. If you already use one of these CLIs on this machine, the Import from CLI setup button pulls your existing keys and sign-ins across with one click — no pasting.
You need at least one connected model for any agent to run. You can add more from Settings → Providers at any time. If you want to be frugal, Ollama gives you a free local option with no cost tracking.
Step 4 — Set a daily budget
Budget is one of the first things to configure. Pick a preset or enter daily and monthly limits in US dollars. When the daily spend crosses the limit, Exolvra stops sending new LLM calls until the clock rolls over.
Recommended defaults for a personal install: $5/day. For a team: $50/day and tune up from there. Per-session warnings are configurable later from Settings → Budget.
Step 5 — Ready
The last step confirms everything is wired up and shows your built-in agent roster. Jump straight into Create a project or Start a chat.
If something goes wrong
The Welcome screen says “already set up” but you don’t know the password. Someone (possibly an earlier you) already created the admin account. Use the password reset below.
The setup wizard won’t save a provider. Usually this means no credential was entered. Go back to the model step and add a real API key, sign in to a CLI, or select Ollama and leave the base URL at the default.
You lost your admin password. Run Exolvra once with the --reset-admin flag — it resets the admin credentials and exits without starting the server; then start it normally. For a container deploy: docker exec <container> dotnet Exolvra.Gateway.dll --reset-admin (or the podman equivalent). If you’re not sure how your instance is launched, ask whoever administers it.
Next
- Your first agent — start using the dashboard
- Admin overview — the post-install hardening checklist