Autonomis

Dashboard Guide

Your command center for managing AI agents, chatting, monitoring usage, and configuring every aspect of the platform. This guide walks through every feature you will find in the dashboard.

Sidebar Navigation

The left sidebar is always visible on desktop. On mobile, tap the hamburger menu to open it. It is organized into everyday tools and advanced sections:

  • Overview — The main dashboard home screen
  • Group Chat — Multi-agent conversation
  • New Agent — Deploy a new agent
  • Templates — Edit agent personality and behavior files
  • Secrets — Manage encrypted environment variables
  • Skills — Browse and install pre-built capabilities
  • Integrations — Connect Discord, Telegram, Slack, WhatsApp, iMessage

Under the Advanced section you will find:

  • Usage — Detailed token consumption and cost estimates
  • Files — Browse agent workspace files
  • Memories — Search your agents' long-term memory
  • Tasks — View and manage background tasks
  • Webhooks — Incoming webhook configuration
  • Cron Jobs — Scheduled, recurring tasks
  • Alerts — Notification rules
  • Logs — Container logs for debugging

At the bottom of the sidebar you will find Settings for account options.

Overview Page

This is the first screen you see after signing in. It gives you a bird's-eye view of everything happening in your workspace.

Stats Cards

Summary cards sit at the top of the page:

  • Agent Count — Shows how many agents you have deployed versus your maximum of 5. For example, "3 / 5" means you are running 3 agents and can add up to 2 more.
  • Total Resources — The combined CPU and memory allocated across all your agents.

Agent Cards

Below the stats, a grid of cards shows each of your deployed agents. Every card includes:

  • Agent name and display name
  • Role — for example Manager, Researcher, or a custom role you defined
  • Status indicator — a small colored dot: green means online, red means offline
  • "Default" badge — marks the orchestrator agent, which is the lead agent that coordinates work among other agents

Click any agent card to open its detail page.

"+ New Agent" Card

When you still have room for more agents (up to 5), a special card appears at the end of the grid. Click it to go straight to the agent creation page. If you have reached the 5 agent limit, this card will not appear.

Usage Summary

A summary section shows your token usage for the current month:

  • Total Tokens — Combined input and output tokens across all agents, displayed in thousands (K) or millions (M).
  • Estimated Cost — Approximate cost based on the AI models your agents used.
  • Per-Agent Breakdown — A table listing each agent's model, token counts, request count, and estimated cost.
Good to know: Autonomis uses BYOK (Bring Your Own Key). The token costs shown here are estimates. Actual AI usage charges appear on your provider's bill (Anthropic, Google, OpenAI, or Moonshot).

Quick Actions Panel

A panel of one-click shortcuts keeps common tasks within easy reach:

  • Group Chat — Jump straight into a conversation with all your agents
  • Secrets — Open the encrypted secrets manager

Agent Detail Page

Click any agent card on the Overview page to open its detail page. This is where you monitor, chat with, and manage a single agent.

Header

At the top you will see the agent's name, role, online/offline status, and the AI model it is currently using (for example, Claude Sonnet or Gemini Pro).

Chat Button

A prominent Chat button lets you open a direct, 1-on-1 conversation with this specific agent. See the Single Agent Chat section below for details.

Resources

Each agent gets its own dedicated resources. The detail page shows exactly what is allocated:

  • CPU — Processing power (200 MHz per agent)
  • Memory — RAM (1.5 GB per agent)

Every agent runs in its own isolated container, so one busy agent cannot slow down another.

Usage Stats

The detail page shows usage statistics for this specific agent: tokens consumed, number of requests, and estimated cost. This helps you understand which agents are doing the most work.

Management Actions

Three management actions are available on the detail page:

  • Restart Agent — Reboots the agent's container. Useful if an agent becomes unresponsive or gets stuck. Your conversation history and files are preserved.
  • Rotate Key — Generates a brand-new API key for the agent. Use this periodically for security, or immediately if you suspect a key has been compromised. The old key stops working right away.
  • Delete — Permanently removes the agent, its container, and its workspace data. This action cannot be undone, so the dashboard will ask you to confirm before proceeding.

Group Chat

Group Chat lets you talk to all your agents at once in a shared conversation. It is the most powerful way to use a multi-agent setup.

When you send a message, your lead agent (the one with the "default" badge) reads it first and decides how to handle it. If the task needs a specialist, the lead agent automatically delegates work to the right agent.

You can also bring a specific agent into the conversation manually by typing [INVOLVE:agent-name] in your message. For example, [INVOLVE:researcher] tells the lead agent to loop in your researcher agent for that task.

Responses stream in real-time, so you see the agent's reply as it is being generated — no waiting for the full response.

Single Agent Chat

For focused work, open a direct 1-on-1 conversation with any agent from its detail page.

  • Conversation history is preserved between sessions. Close the browser, come back later, and your conversation picks up right where you left it.
  • Vector memory means your agent remembers context from previous chats. It automatically stores important information and can recall it when relevant.
  • Delegation works here too. Even in a 1-on-1 chat, the agent can delegate tasks to other agents if needed using the [INVOLVE:agent-name] pattern.

Cron Jobs (Scheduled Tasks)

Cron Jobs let you set up recurring tasks that run automatically on a schedule — no need to be online. Great for daily summaries, periodic research, automated reports, or regular check-ins.

Creating a Cron Job

To set up a new scheduled task:

  1. Go to Cron Jobs in the sidebar under Advanced.
  2. Click New Cron Job.
  3. Fill in the details:
    • Name — A descriptive label (e.g., "Daily news summary")
    • Schedule — When it should run. You can use cron format (e.g., 0 8 * * *) or simple expressions like "every day at 8am"
    • Command — The instruction to send to the agent
    • Model — Which AI model to use for the task
    • Timezone — Your local timezone so schedules run at the right time
  4. Click Save.

Managing Cron Jobs

Once created, you can enable or disable individual jobs with a toggle, edit the schedule or command, or delete jobs you no longer need. The execution history shows when each job last ran and whether it succeeded.

Templates

Templates are simple text files that shape how your agents think, communicate, and behave. You can edit them directly in the dashboard — no coding required.

The Four Template Files

  • SOUL.md — The agent's core personality and identity. Define who the agent is, how it should communicate, and any behavioral rules it should follow.
  • USER.md — Information about you. Tell the agent your name, your preferences, the kind of projects you work on, and anything else it should know about you.
  • TOOLS.md — Instructions for using tools and skills. This tells the agent how to use any special capabilities it has access to.
  • HEARTBEAT.md — Periodic reminders that run every few minutes. Use this for status checks, recurring instructions, or anything the agent should keep in mind continuously.

User-Level vs. Per-Agent Templates

Templates work on two levels:

  • User-level templates apply to all your agents by default. Set these first to establish a baseline personality and context.
  • Per-agent overrides replace the user-level template for a specific agent. For example, you might give your researcher agent a different SOUL.md that makes it more analytical, while your writer agent gets one that makes it more creative.

To edit user-level templates, go to Templates in the sidebar. To set per-agent overrides, go to the agent's detail page and look for the Templates section.

Secrets

Secrets let you securely store sensitive data that your agents need — things like API keys for external services, database passwords, or authentication tokens.

  • All secrets are encrypted using Vault Transit encryption. Even platform administrators cannot see the plaintext values.
  • In the dashboard, secret values are always masked. You can see the key names but not the values.
  • Secrets are automatically available as environment variables inside your agent containers. Your agents can use them without any extra configuration.

Managing Secrets

  1. Go to Secrets in the sidebar.
  2. Click Add Secret, enter a key name (e.g., GITHUB_TOKEN) and its value.
  3. Click Save. The secret is encrypted and stored.
  4. To remove a secret, click the delete icon next to it.

Skills Marketplace

Skills are pre-built capabilities that add new tools and abilities to your agents. Think of them as plugins that extend what your agents can do.

  • Browse the marketplace to discover available skills.
  • Install a skill with one click to make it available to your agents.
  • Manage your installed skills — enable, disable, or remove them as needed.

Integrations

Connect your agents to messaging platforms so you can interact with them wherever you already communicate. Available integrations:

  • Discord
  • Telegram
  • Slack
  • WhatsApp
  • iMessage

Each integration includes a step-by-step setup guide right in the dashboard. Follow the instructions, connect your account, and your agents will be reachable from that platform.

Usage Tracking

The Usage page (under Advanced in the sidebar) gives you a detailed view of how your agents are consuming tokens.

  • Per-agent breakdown showing the model used, input tokens, output tokens, number of requests, and estimated cost.
  • Monthly totals across all agents.
Reminder: Autonomis uses BYOK (Bring Your Own Key). The costs shown in the dashboard are estimates to help you track spending. Actual AI usage charges appear on your provider's bill (Anthropic, Google, OpenAI, or Moonshot).

Files

The Files page lets you browse your agent's workspace — the directory where your agents create, read, and store files during their work.

Each agent has its own workspace directory. You can navigate through folders and view file contents directly in the dashboard. This is useful for reviewing reports, scripts, notes, or any other files your agents have generated.

Memories

Your agents automatically build a long-term semantic memory from their work and conversations. The Memories page lets you search and browse what your agents have remembered.

  • Automatic — Agents store important information on their own. You do not need to tell them what to remember.
  • Searchable — Type keywords to find specific memories.
  • 90-day retention — Memories are kept for 90 days and then automatically cleaned up.

This is what allows your agents to recall context from previous conversations and tasks, making them more helpful over time.

Tasks

The Tasks page shows background tasks that your agents are working on. You can view the status of each task, see progress, and manage them as needed. This is particularly useful for long-running operations that you do not want to wait on in a chat window.

Logs

The Logs page shows the raw output from your agent containers. This is primarily a debugging tool — if an agent is not behaving as expected, checking the logs can help you (or the support team) understand what went wrong.

You do not need to check logs during normal use. They are there when you need them.

Settings

The Settings page is where you manage your account:

  • Account — Update account details and preferences.

Next Steps

Now that you know your way around the dashboard, here are some good places to go next: