agentos.guide › blog
Hermes Agent Bot Mode: What It Is, A Five-Minute Quick-Start And How It Fits An Agent OS
Hermes Agent Bot Mode is a free feature of the Hermes desktop app that turns your Hermes profiles into a roster of named bots, each with its own chat, model, memory, skills, avatar and schedule.
Nous Research made it, it is MIT licensed, and the official docs say it now ships built into the desktop app and is on by default.
Below you'll get what it is, a five-minute quick-start, how to create a bot, every feature, the Agent Inbox, the limits, my Grok Bot comparison and how it fits an Agent OS.
Hermes Agent Bot Mode at a glance
Here is the whole answer in one table before the detail.
| Question | Short answer |
|---|---|
| What is it? | It is a roster of named bots built on top of Hermes profiles. |
| Who made it? | Nous Research made it, and Teknium shipped the first beta. |
| What does it cost? | The software is free and MIT licensed, and model usage is the only cost. |
| Where does it live? | It lives in the Hermes desktop app as a Bots tab next to Sessions. |
| How do I install it? | The docs say it ships built in and is on by default. |
| What does each bot get? | Each bot gets its own chat, model, memory, skills, avatar and routines. |
| Can bots talk to each other? | Bots message each other with attribution and can share group chats. |
I installed Bot Mode on launch day and documented that real install in my original Hermes Bot Mode guide.
Everything newer than that build comes from the official Bot Mode docs, which I re-read on 12 October 2026.
I'll tell you which facts come from my own install and which come from the docs.
What Hermes Agent Bot Mode is and the problem it solves
Hermes Agent is the free, open-source AI agent from Nous Research.
A Hermes profile is one isolated agent setup with its own config, memory, skills, credentials and chat history.
Bot Mode shows each profile as a named bot in a roster down the left of the desktop app.
You click a bot and you land in its chat, the same way you'd message a person on your team.
The docs say a bot is a profile, so there is no new primitive to learn.
Before Bot Mode, my inbox summary, my content research and my competitor tracking lived in three sessions that looked identical.
I was the routing layer between them.
A roster gives every job a name, a face and its own memory.
Who made Bot Mode and what it costs
Nous Research made Bot Mode.
Teknium from the Nous team shipped it as a public beta plugin in the NousResearch/Hermes-Bot-Mode repository under an MIT licence.
GitHub shows that repository was created on 13 August 2026.
Somebody said it looked a lot like Grok Bot, and Teknium answered, "That's the point."
The repository is now archived, and its README says Bot Mode became a bundled, default-on desktop plugin through pull request #87886.
GitHub shows that pull request was merged on 17 August 2026.
The software is free.
You pay only for the model behind each bot, and a bot on a local model has no model bill.
The five-minute quick-start
- First, install Hermes Agent with the official script if you haven't already.
curl -fsSL https://hermes-agent.nousresearch.com/install.sh | bash- Second, run
hermes setupand connect a model provider. - Third, run
hermes updateso you're on a current build. - Fourth, open the desktop app, which the docs say you can launch with
hermes desktop. - Fifth, click the Bots tab next to Sessions in the left sidebar.
- Sixth, press New Agent and type a Name, a Title and a Description.
- Seventh, open the Routines tile and give that bot one recurring job.
The docs say the desktop app runs on macOS with Apple Silicon, Windows and Linux.
Your existing profiles appear as bots the moment you open the tab, so you rebuild nothing.
Older tutorials tell you to clone the plugin into a desktop-plugins folder first.
The archived README says no manual install is needed on current desktop builds.
How to create a bot in Bot Mode
Creating a bot took me about twenty seconds in my guide.
I typed "scout", then "Competitor scout", then one line about its job.
A real profile appeared on disk with its own memory, skills and inbox.
The docs say a new bot introduces itself as the first message of its own chat.
The Advanced section is the part most people scroll past.
| Advanced option | What it does |
|---|---|
| Clone from | It starts the bot from another profile's config, skills, SOUL and memory. |
| Create empty | It skips the bundled skills for a minimal profile. |
| Model and provider pin | It gives this bot its own model, so different bots run different models. |
| Custom SOUL.md | It sets the bot's persona and standing instructions. |
| Skills, toolsets and MCP servers | It lets you tick exactly what this specialist can use. |
| Copy API keys | It copies static API keys from the main profile and is on by default. |
| Create on | It picks which connected machine the bot lives on. |
The model pin is the setting I called out in my video.
My research bot runs one model, my writing bot runs another and my inbox bot runs a small local model.
The docs flag one catch, which is that single-use OAuth logins for Anthropic, OpenAI Codex and xAI are not copied to a new bot.
You sign that bot in with hermes -p <name> auth add <provider>.
🔥 Want the exact Hermes setup I run inside my Agent OS?
Inside the AI Profit Boardroom you get the installable Agent OS, step-by-step video tutorials, four coaching calls a week and 3,400+ members building real automations.
Every Hermes Agent Bot Mode feature
| Feature | How you use it | Source |
|---|---|---|
| Bots pane | You click a row to open that bot's chat. | My guide and the docs |
| Active now | You filter the roster to the bots working right now. | The docs |
| Bot Chat | You return to one permanent conversation per bot. | The docs |
| Routines | You attach a recurring job to the bot that does it. | My guide and the docs |
| @mentions | You type @name in any chat to hand work off. | My guide and the docs |
| Bot-to-bot messages | A bot messages a teammate with its name attached. | My guide and the docs |
| Group chats | You open a shared room for two to six bots. | The docs |
| Duplicate | You right-click a bot to clone it with its memory. | My guide and the docs |
| Sections and hide | You file bots into folders or hide the ones you don't use. | The docs |
| Avatars and pets | You pick a face, an image, a portrait or a pixel pet. | My guide and the docs |
| Voices | Each bot speaks with its own text-to-speech settings. | The docs |
| Bots across machines | You see bots from every machine you connect. | The docs |
The Bots pane and Bot Chat
The roster shows one row per profile with an avatar, a message preview and a timestamp.
Every bot has one canonical conversation called Bot Chat.
Typing /new in that chat compacts the context instead of forking a new session.
You can still open a bot's other sessions from its right-click menu.
Routines
A routine is a recurring task attached to the bot that does it.
In my guide I gave my inbox bot an "Evening wrap-up" routine at 6pm, and Hermes cron returned the next run time.
Routines are plain cron jobs named [bot:<name>] <routine>, so they also show in hermes cron list.
The result lands in that bot's own chat, so it is waiting when you click the bot.
Memory, SOUL.md and cloning
Each bot keeps its own memory, and Teknium confirmed that at launch.
Every bot has a SOUL.md file that defines how it talks, what it cares about and what it never does.
Duplicate makes a full clone, including config, skills, SOUL.md, memory and the bot's look.
A research bot you've trained for weeks can become the starting point for a new project.
Avatars, pets and voices
The docs list blob faces drawn from the bot's name, geometric faces in seven shapes and ten colours, uploaded images, AI portraits and pixel pets.
A pet bounces beside the avatar while the bot is busy.
With six bots on the roster and three mid-task, you can see who is working at a glance.
The docs also say each bot speaks with its own profile's text-to-speech settings.
Group chats, sections and bots across machines
These three arrived after my launch-day guide, so this part is from the docs.
A group chat is a shared room for two to six bots.
Your message triggers up to three rounds of turns, and each bot replies briefly or passes.
A bot can escalate to you with @user, and the room shows a "needs you" badge.
Sections are folders you name yourself, such as Clients or Team.
The roster can also show bots from every backend you connect, including SSH hosts and Hermes Cloud.
The Agent Inbox and bot-to-bot messaging
The Agent Inbox was the launch-day name for a bot's persistent conversation where messages from other agents land.
The current docs call each bot's permanent conversation Bot Chat, and agent messages are delivered into it.
Every message arrives with attribution, so the receiving bot knows exactly which agent sent it.
The docs show the format as "Message from 🤖 researcher (@researcher):".
You can trigger a handoff yourself with an @mention.
- You type
@researcher have a look at thisin any bot's chat. - The active bot composes its own message to the researcher.
- The researcher runs with its own memory and its own model.
- The answer comes back in the same conversation, attributed to the researcher.
I showed this live in my guide by asking my writer bot to pull one line from my researcher bot.
The docs say bots send these messages with a tool called message_agent, and your text is never forwarded word for word.
Delivery is fire-and-forget, so the sender finishes its turn and the reply arrives as a notification.
An early tester named Devin Davis said he spent a week trying to coordinate two profiles with a kanban board.
He had bot-to-bot coordination running the same morning he tried Bot Mode.
On my launch-day install, a bot's very first inbox message failed because the inbox thread didn't exist yet.
The current docs describe a flag that creates the conversation when it is missing, which removes that problem.
🔥 Want the exact Hermes setup I run inside my Agent OS?
Inside the AI Profit Boardroom you get the installable Agent OS, step-by-step video tutorials, four coaching calls a week and 3,400+ members building real automations.
How I wired Bot Mode into my Agent OS
My Agent Operating System is one dashboard for every agent I run.
I rebuilt the Bot Mode roster inside it as a Bots tab in the Hermes section.
| Layer in my Agent OS | What it does | Where Bot Mode fits |
|---|---|---|
| The dashboard | One screen shows every agent I run. | The Bots tab sits in the Hermes section. |
| The profiles | Each Hermes profile is one isolated agent. | Each profile appears as one bot. |
| The schedule | Hermes cron runs the recurring jobs. | A routine is a cron job named after its bot. |
| The handoffs | Agents pass work to each other. | A bot messages a teammate with attribution. |
| The shared memory | My Obsidian vault holds what every agent should know. | Each bot still keeps its own private memory. |
The tab reads the same Hermes profiles the desktop app reads, so there is one team and one source of truth.
My guide shows 46 bots in that roster, and every one is a real profile on my machine.
Oracle watches my competitors, Astros pulls keyword ideas out of search data and Apollo handles voice.
Those used to be separate profiles I hopped between.
Now one bot messages another without me copying and pasting between windows.
The design is easy to build on because a bot is only a profile and a routine is only a cron job.
My Agent OS build guide shows how the dashboard itself goes together.
My Hermes Command Desk guide shows how Hermes sits in every room of that dashboard.
What changed since launch day
My original guide documents a one-day beta, and the feature has moved since.
| Topic | Launch day in my guide (14 Aug 2026) | Current docs (read 12 Oct 2026) |
|---|---|---|
| Install | You cloned a plugin folder from GitHub. | It ships built in and is on by default. |
| The bot's main thread | It was called the Agent Inbox. | It is called Bot Chat. |
| Group chats | The answer was "not yet". | Rooms for two to six bots are documented. |
| Deleting a bot | It was command-line only, on purpose. | Delete Profile is in the right-click menu behind a confirmation. |
| Live interrupt | It was on the future list. | It is still listed as future work. |
I haven't re-documented the newer features in my own guide yet, so the right-hand column is the docs talking.
Limits to know before you build a team
- Bot Mode lives in the desktop app, so a headless server on its own won't show a Bots tab.
- The docs say the
message_agenttool doesn't appear on a headless install until you add two markers by hand. - Bot-to-bot delivery is per-invocation, so the receiving bot reads a message when it next runs.
- Live interrupt of a bot in the middle of a conversation is listed as future work.
- The desktop keeps three local bot backends warm by default, and the docs put each at about 60 MB.
- A bot opened while every slot is busy waits up to 30 seconds, so raise Warm Bot Backends for a large team.
- A bot with no routine sits idle, and one of my viewers noticed bots going quiet for that reason.
- You can switch Bot Mode off in the desktop app's plugin settings, and the docs say your profiles, sessions and cron jobs stay untouched.
A viewer told me the interface looked a bit ugly next to Grok Bot at launch, and that was fair for a one-day beta.
Hermes Agent Bot Mode vs Grok Bot
This is the comparison from my video and my Hermes Bot vs Grok Bot post, which I made in August 2026.
| What I compared | Hermes Agent Bot Mode | Grok Bot |
|---|---|---|
| Software cost | It is free and MIT licensed. | It is a paid platform. |
| Model per bot | Any model works, and each bot can differ. | It runs on xAI models only. |
| Where it runs | It runs on your own computer. | It runs in xAI's cloud. |
| Bot memory | It stays in folders you own. | It sits on their servers. |
| Code | You can read it. | You can't read it. |
| Interface at launch | It was rough, as a one-day beta. | It was smoother that day. |
Grok Bot isn't bad, and it showed the world what a team of named agents looks like.
Bot Mode gives you the same shape with different ownership.
One tester in the community had Hermes rebuild their whole Grok Bot team, coordinator and workers, inside Bot Mode.
Grok Bot may have changed since August, so check its current page before you choose.
My Grok Bot course covers that product on its own terms.
Related guides in this batch
My Hermes Agent bot guide covers Hermes as a bot on Telegram, Discord and Slack.
My Hermes Agent background guide covers cron, background tasks and the gateway service.
My best Hermes Agent UI ranking shows where the desktop app sits against other front ends.
Also On Our Network
FAQ
What is Hermes Agent Bot Mode?
Hermes Agent Bot Mode is a feature of the Hermes desktop app that shows your Hermes profiles as a roster of named bots. Each bot has its own chat, model, memory, skills, avatar and routines, and bots can message each other.
Who made Hermes Agent Bot Mode?
Nous Research made it. It launched in August 2026 as a public beta plugin in the NousResearch/Hermes-Bot-Mode repository under an MIT licence, and that repository is now archived because Bot Mode ships built into Hermes Desktop.
Is Hermes Agent Bot Mode free?
Yes. The software is free and MIT licensed. Your only cost is the model usage behind each bot, and a bot pinned to a local model has no model bill.
How do I install Hermes Agent Bot Mode?
The official docs say it ships built into the Hermes desktop app and is on by default. Update Hermes, open the desktop app and click the Bots tab next to Sessions.
What is the Agent Inbox in Bot Mode?
The Agent Inbox was the launch-day name for a bot's persistent conversation where messages from other agents arrive with attribution. The current docs call each bot's permanent conversation Bot Chat.
How does Bot Mode fit into an Agent OS?
A bot is a Hermes profile and a routine is a Hermes cron job, so a dashboard can read the same roster. I added a Bots tab to the Hermes section of my Agent OS that shows the same bots, chats and routines.
Final word on Hermes Agent Bot Mode
Update Hermes, open the Bots tab and create one bot with one job and one routine.
Check its chat tomorrow morning, and you'll know whether your work is ready for a full team in Hermes Agent Bot Mode.
About Julian
I'm Julian Goldie, an AI entrepreneur, SEO expert and the founder of the AI Profit Boardroom, which has 3,400+ members.
I help business owners scale with AI agents, automation and SEO.
My YouTube channel has 400,000+ subscribers, and I run Goldie Agency, a seven-figure SEO agency.
Related guides on agentos.guide
→ Hermes Bot Mode: The Named Agent Engine
→ How to Build Your Own AI Agent OS
Get the whole Agent OS
The installable system behind every guide on this site — dashboard, agents, memory and pipelines — updated as new tools land, with 3,400+ members and four coaching calls a week.
Join the AI Profit Boardroom →📺 Video notes + links to the tools 👉
🎥 Learn how I make these videos 👉