Auxot
Tutorials Docs Pricing Login

Find the right screen for your next question

Open each primary destination once, learn what kind of work belongs in **Chat** versus **System Health** versus **Context Files**, **Workflows**, and **Settings**, and stop opening the wrong drawer when you have a question.

Plus: three Admin Agent prompts that turn the left menu from wallpaper into a labeled map tied to your org, your role, and what is already wired.

Audience Everyone
Time ~5 min
Prerequisites You reach the main Auxot shell without **onboarding-blocked**. Owners: [Finish first-run onboarding](/tutorials/finish-first-run-onboarding). Everyone else: helpful if you already skimmed [Unstick your first week in Auxot](/tutorials/unstick-your-first-week-in-auxot) when something felt stuck.
You'll end up with A five-minute mental map of the authenticated shell: which left-menu item answers which class of question, plus three Admin Agent prompts you can reuse when the product grows new pages.

When a tutorial shows italic text in quotation marks, it usually mirrors a label or helper string inside Auxot. Product copy changes between releases — if something reads differently in your workspace, trust what you see on screen.

Callouts with a Worth knowing gold accent are meant as must-read context before you move on. Blockquotes that open with Tip are lighter, optional depth.

Why this matters

Auxot packs a lot behind one sign-in. Chat is not Settings. System Health is not Audit Logs. Context Files are not the same thing as a long chat thread. When every surface is new, people default to clicking at random, filing fake bugs, or asking the Admin Agent for help without naming what they are looking at.

Your left menu is a list of pages, each with a specific job. Learn what each page is for once, and every later tutorial (models, Slack, workflows, keys) lands faster because you already know where it takes you.

Today, you open each primary destination once and attach a plain-English job description to it. The next time you think something is wrong, you will know whether to start in Chat, System Health, or Settings before you burn an afternoon.

The map is about labels on screen, not hidden infrastructure. When a card names a database or a worker, you still read it in System Health first.

Quick start

  1. Confirm the shell: sign in and land on /app without the onboarding-blocked warning. If you cannot, stop here and use Unstick your first week in Auxot or Finish first-run onboarding.
  2. Open Chat: click Chat in the left menu. This is where you talk to the Admin Agent and any custom agents (Say hello to the Admin Agent).
  3. Open System Health: click System Health. This is where services, providers, workers, and recent activity show whether the platform is healthy (Take Auxot’s pulse in 10 seconds).
  4. Open Context Files: click Context Files. This is where durable knowledge for agents lives, scoped org or user (Add your first context file).
  5. Skim the rest once: click Dashboards, Workflows, Audit Logs, and Profile in order, then open Settings and read whichever sub-pages you actually see (labels vary by role and license).

Done? You can finish this sentence for each item you opened: When my question is about ____, I start in ____.

The agent can do that?

You have clicked the shell once. These three prompts ask the Admin Agent to label the map for your account (orientation and configuration language, not business execution).

1. Turn the left menu into a lookup table

I just finished a five-minute click-through of the main left menu (Chat, System Health, Dashboards, Context Files, Workflows, Audit Logs, Profile, Settings). In a markdown table, one row per item I listed: what kind of question belongs there, what kind of question does *not* belong there, and one concrete example for each row. Use only Auxot product vocabulary from the UI labels, not generic AI advice.

Why it’s non-obvious: New users treat every screen as the AI. Once you separate conversation from configuration from evidence of what ran, the rest of the product gets a lot easier to read.

2. Route a single real symptom to the right first screen

Symptom in one sentence: [paste what you see or what failed]. Tell me the *first* place I should click in the left menu (exact label), what I should verify on that page in under thirty seconds, and one follow-on screen only if the first check passes. If the answer is *wait for an admin*, say that plainly.

Why it’s non-obvious: People open Chat for outages and Settings for model behavior. Naming the first place to look stops the panic-click spiral before it starts.

3. Inventory what your account already has wired

Assume I am a new member who can read but not change admin settings. From what you can infer about this org, list what is probably already configured (providers, agents, integrations) versus what is probably still missing. Keep it to eight bullets, each starting with the left-menu area I would check to confirm (Chat, System Health, Context Files, Workflows, Audit Logs, Profile, Settings).

Why it’s non-obvious: The empty state and the misconfigured state look alike until you know where evidence lives.

Go deeper

Chat versus Settings (the short version)

Chat is for prompting agents in language: questions, drafts you iterate on with a persona, and asking the Admin Agent how Auxot works.

Settings is for changing how the org behaves: agents, providers, users, keys, credentials, tool policies, and license surfaces you are allowed to see. If the fix sounds like flip a switch for everyone, you are closer to Settings than to Chat.

System Health versus Audit Logs

System Health answers is the platform and its integrations healthy right now, with cards you can read in seconds.

Audit Logs answers what jobs or events happened, searchable history for forensics and habits (View your audit logs). Red health and a quiet audit trail still belong to different investigations.

Dashboards, Workflows, and Context Files (how they stack)

Dashboards aggregate signals for boards you or your team set up (Build your first Dashboard).

Workflows run multi-step automation you or your team designed (Run a workflow).

Context Files hold text your agents reuse across conversations. They do not replace workflows, and they do not replace dashboards.

Profile, Escalations, and admin-only Settings

Profile is your identity, password, theme, and sessions (Customize your Profile).

Escalations (bell panel) is where human review requests surface when a workflow or tool policy demands it (Set up an Escalation).

Some Settings sub-pages are admin-only (for example Settings → Providers or Settings → Users). If you cannot see a page a tutorial mentions, you are not blocked: you still learned that the org routes that work through an admin.

Walkthrough

Step 1: Confirm you are in the right org and past the gates

Glance at the org switcher or profile corner so you are not mapping Chat for Org A while your teammates live in Org B. If onboarding-blocked still appears, the shell tour can wait.

Step 2: Chat (conversations, not switches)

Open Chat. Pick the Admin Agent if the picker is available. Send nothing yet if you prefer. The goal is to notice this is where threads live. Custom agents appear in the same picker after your team creates them.

Step 3: System Health (truth about now)

Open System Health. Scroll once. You are learning the vocabulary cards use (providers, workers, timestamps), not memorizing every line item.

Step 4: Context Files (durable memory)

Open Context Files. Notice scope language (org versus user) and titles. You are not writing a file yet unless you want to; you are anchoring long-lived knowledge attaches here.

Step 5: Sweep the remaining primary destinations

Open Dashboards, Workflows, Audit Logs, and Profile once each, even if some are empty. Finish with Settings: click every sub-page you can see, read the page title, and stop. Optional: return to Chat and ask the Admin Agent for the table prompt from The agent can do that? while the UI is still fresh.

What’s next

Reference

Browse all tutorials