Hermes advanced delegation is the official Nous Research plugin that adds two new ways to route subagent work in Hermes Agent: model routes and profile routes.
It plugs into the delegate_task tool, so the main agent can choose, per task, which model or which Hermes profile a subagent runs on.
Without it, Hermes gives every subagent in a batch the same model.
The plugin is free, it is part of the Hermes Official Plugins, and it needs Hermes 0.22.5 or newer.
This page is the operator setup in five steps, from install to your first routed batch, using only commands from Nous's own plugin page and docs.
The Hermes Advanced Delegation Setup At A Glance
Here is the full setup on one screen.
| Step | What you do | Command or setting |
|---|---|---|
| 1 | Update Hermes to 0.22.5 or newer. | hermes update |
| 2 | Install the plugin in the profile that delegates. | hermes plugins install advanced-delegation |
| 3 | Pick a mode and build your routes. | hermes delegation config |
| 4 | Check what you saved without opening an editor. | hermes delegation config --show |
| 5 | Switch modes from a chat and start a fresh session. | /delegation, then /new |
The rest of this walkthrough explains each step, then the config file version for anyone who prefers YAML.
What Hermes Advanced Delegation Actually Does
Hermes already has subagents built in, through the delegate_task tool.
With the built-in default, every subagent runs on delegation.model and delegation.provider, or on the parent's model if those are empty.
The plugin page lists the three modes side by side.
| Mode | delegation.backend |
What the agent gets |
|---|---|---|
| Default, built into Hermes | default |
Every subagent runs on the delegation model, or the parent's model. |
| Model routes | model_map |
A per-task choice of a named model route from a list you define. |
| Profile routes | profile_routes |
A per-task choice of one of your Hermes profiles, with its own model, credentials, skills and SOUL.md. |
Only one mode is active at a time.
Model routes and profile routes keep separate config blocks, so switching modes keeps the other mode's settings for when you switch back.
The plugin page says Hermes core still handles the whole subagent lifecycle, and the plugin only answers which route a task takes.
If you want the basics of subagents first, read my sibling post on Hermes Agent subagents.
🔥 Want my exact Hermes Advanced Delegation setup, step by step? Inside the AI Profit Boardroom you get step-by-step video trainings, my installable Agent OS and four coaching calls a week with 3,400+ members building real automations. → Join the AI Profit Boardroom
Step 1: Update Hermes To 0.22.5 Or Newer
The plugin needs Hermes 0.22.5, which is the first release with pluggable delegation providers.
The v0.22.5 release on GitHub, published on 23 October 2026, lists pluggable delegation backends that route subagents by model map or profile.
The plugin's manifest declares requires_hermes: ">=0.22.5".
The plugin page says Hermes 0.22.0 to 0.22.4 refuses to install or enable it, and older cores load it but it does nothing.
Run the update, then check your version.
hermes update
hermes --version
Step 2: Install The Plugin In The Right Profile
The plugin page gives one install command from the Hermes plugin catalogue, which is pinned to a reviewed commit.
hermes plugins install advanced-delegation
Desktop users can open Settings → Model → Subagents and click Install plugin, which shows under the mode picker while the plugin is missing.
If you want to follow the plugin repo's main branch instead of the catalogue pin, the page gives this command.
hermes plugins install NousResearch/hermes-official-plugins#advanced-delegation --enable
Plugins install per profile, so install it in the profile that does the delegating.
Profiles that you use as profile routes need nothing installed, because they only have to exist with a working model.
hermes profile list shows which profiles you have.
Step 3: Pick A Mode And Build Your Routes
The plugin adds a guided setup command.
hermes delegation config # pick the mode, then edit its routes
hermes delegation config --show # print the current mode and routes (no terminal needed)
First you choose Default, Model routes or Profile routes, and the cursor opens on your current mode.
For model routes, you get a route editor built from hermes model's own pickers.
Each route needs a name, which is what the agent picks, and a one-line description, which is what the agent reads.
Then you choose a provider you already have credentials for, or a custom endpoint, and a model from the picker.
You can add an optional reasoning effort and a speed tier when the route offers one.
The plugin validates each route before saving, so the provider must be known, the effort must be a real level and the tier must be one the route offers.
When you finish, it asks you for a default route, or you can choose no default so unrouted tasks use the main agent's model.
For profile routes, the editor is simply a checklist of your other profiles, an optional description for each and a default.
Profiles keep their own models, so there is nothing else to pick.
Everything is written to the profile's config.yaml with Hermes's own config writer, which keeps your comments and other settings.
🔥 Want my Hermes delegation setups done with you, step by step? Inside the AI Profit Boardroom, I've got a full Hermes Agent course plus the Agent OS, with video walkthroughs on profiles, models and subagents. You'll be learning alongside 3,400+ members with four coaching calls every week. → Get access here
Step 4: Or Configure It By Hand
If you prefer YAML, each mode's settings live under delegation.backends.<mode>, and delegation.backend picks the active mode.
This is the model routes example from the plugin page.
delegation:
backend: model_map
backends:
model_map:
options:
- name: fast # what the agent picks
description: "Cheap and quick: lookups, summaries, simple edits."
provider: openrouter # provider and/or base_url, plus model
model: google/gemini-3-flash-preview
- name: deep
description: "Slow and careful: hard debugging and design work."
provider: anthropic
model: claude-opus-5
reasoning_effort: high
service_tier: fast # optional speed tier: normal, fast or ultrafast
- name: local
description: "Private: runs on this machine."
base_url: http://localhost:8080/v1
model: qwen3-coder
default: fast # optional: route for tasks that pick none
Each route needs name, description and model, plus provider or base_url.
Optional keys are reasoning_effort, service_tier, fallback_providers, api_key and api_mode, each meaning the same as the matching delegation.* key.
The plugin page recommends provider: over api_key:, because an api_key here is stored in config.yaml in plain text.
This is the profile routes example from the same page.
delegation:
backend: profile_routes
backends:
profile_routes:
profiles: # allowlist AND what the agent reads about each profile
- name: coder
description: Writes and refactors code; runs tests.
- name: researcher
description: Web research and long-form summaries.
default: coder # optional: profile for tasks that pick none
The list doubles as an allowlist, so profiles you don't list can never be picked.
The page also gives command-line versions, such as hermes config set delegation.backend model_map and hermes config set delegation.backends.model_map.default fast.
🚀 Ready to put Hermes Advanced Delegation to work in your business? Inside the AI Profit Boardroom you get step-by-step video trainings, my installable Agent OS and four coaching calls a week with 3,400+ members building real automations. → Join the AI Profit Boardroom
Step 5: Switch Modes In A Chat
The plugin adds a /delegation slash command.
/delegation # current mode, what each mode routes to, a picker to switch
/delegation model_map # save it; it takes effect from the next session (/new)
/delegation profiles --now # save it AND apply it to this chat now
/delegation default # back to the built-in behaviour (from the next session)
Switching saves the mode as the profile's delegation.backend, and it applies from the next session.
The reason is cost, because the tool list is part of the prompt cache and changing it mid-chat makes the next request re-read the whole conversation at full price.
--now is the opt-in for when you want the change in the current chat anyway, at the cost of one prompt-cache break.
On Telegram, Discord and Slack, bare /delegation shows buttons, while the CLI, TUI and Desktop use their own picker.
How A Routed Task Behaves
Once a mode is active, each task item in delegate_task gains one extra property: option for model routes or profile for profile routes.
That property is an enum of your route names in config order, with your descriptions attached.
Hermes refuses any other value before a subagent starts.
A task that picks no route runs on the mode's default, or on the parent's model if there's no default.
While either mode is active, delegation.model and delegation.provider aren't used, and switching back to Default makes them apply again.
Each result reports the route it ran on, including the provider, model, reasoning effort and speed tier, and fallback_activated: true when a fallback took over.
Failures And Fallbacks
The delegation docs say failures are explicit and never a silent switch.
If the selected backend is missing or disabled, delegate_task returns a tool error naming the fix, and no subagent starts.
Inside a batch, a task that can't be routed is refused on its own with route.refused: true, while the other tasks still run.
A model route without its own fallback_providers uses delegation.fallback_providers, and fallback_providers: [] on a route switches fallback off for it.
A profile route uses that profile's own fallback chain.
The Security Warning You Must Read
The plugin page puts this in a warning box, and I agree with it completely.
Listing a profile in profile routes lets this profile's subagents use that profile's credentials, credential pool, SOUL.md and terminal policy.
Only list profiles whose keys and terminal access you're happy for this agent to use.
If any profile runs shell commands, put it behind a sandboxed backend, and my Hermes Agent sandbox walkthrough shows how.
Model Routes Or Profile Routes?
| Question | Model routes | Profile routes |
|---|---|---|
| What changes per task? | The provider and model. | The whole profile, including model, credentials, skills and SOUL.md. |
| What do you set up? | A list of named routes in one profile. | Separate Hermes profiles that already work. |
| Best for | Mixing cheap and expensive models in one batch. | Handing tasks to specialist agents with their own skills. |
| Fallback | The route's own chain, or delegation.fallback_providers. |
The target profile's own chain. |
I'd start with model routes, because they need nothing beyond the one profile you already use.
I'd move to profile routes once you have specialist profiles worth sharing, such as a coder and a researcher.
For choosing the models themselves, my Hermes subagent model walkthrough and best Hermes Agent model guide cover the options.
If you'd like a second pair of eyes on how agents fit your SEO business, you can book a free SEO strategy session with my team.
🔥 Want help designing routes for your own agents? The AI Profit Boardroom has Hermes Agent training, the Agent OS and four weekly coaching calls where you can ask me how to split your work across models and profiles. → Join the AI Profit Boardroom
My video on the Hermes 0.22 release covers the release line this plugin builds on, and my video at the top shows the live subagent controls you'll use to watch any batch, routed or not.
Related Reading
- Hermes Agent subagents explains how
delegate_taskworks underneath the plugin. - Hermes subagent model covers the built-in single model pin.
- Hermes Agent v0.22 covers the release line the plugin builds on.
- Hermes Agent MCP server shows how to give routed agents more tools.
Also On Our Network
- 🌐 How a business would use Hermes advanced delegation
- 🌐 My verdict on model routes versus profile routes
- 🌐 Is Hermes advanced delegation worth it in an Agent OS?
- 🌐 Which models to route to, scored on GoldieBench
- 🌐 My Hermes kanban setup on the AI Profit Boardroom blog
FAQ: Hermes Advanced Delegation
What is Hermes advanced delegation?
It is an official Nous Research plugin that adds model routes and profile routes to delegate_task.
How do I install the Hermes advanced delegation plugin?
Run hermes plugins install advanced-delegation on Hermes 0.22.5 or newer, or use the Install plugin button in Desktop's Subagents settings.
What is the difference between model routes and profile routes?
A model route picks a provider and model, while a profile route runs the subagent as a whole Hermes profile.
Can the agent see my API keys with model routes?
No, because only route names and descriptions reach the model.
Why doesn't my route change apply straight away?
Changes apply to new sessions because the tool list is cached, so use /new or /delegation <mode> --now.
If you run an agency and want help mapping agents to your client work, you can book a free strategy session and we'll talk it through.
📺 Video notes + links to the tools 👉
🎥 Learn how I make these videos 👉
🆓 Get a FREE AI Course + Community + 1,000 AI Agents 👉
About Julian
I'm Julian Goldie, an SEO entrepreneur, author and founder of the AI Profit Boardroom, which has 3,400+ members.
I help business owners scale with AI agents, automation and SEO.
- I've built a 7-figure SEO and link-building agency, Goldie Agency, with 70+ people in Bangkok.
- I've grown a YouTube channel to 400,000+ subscribers.
- I wrote the book "Link Building Mastery", which covers nine link building strategies.
- I run GoldieBench, where I score AI models on real one-shot builds.
→ Get my best AI training inside the AI Profit Boardroom
Update to 0.22.5, install the plugin, build two or three routes and start a fresh session, and Hermes advanced delegation will send each task to the right brain.
