Get free $30 in credits on usBrain is now live for personal and enterprise use. Join leading teams transforming with AI.Join the waitlist

A governed brain for Hermes, past the MEMORY.md ceiling

Hermes Agent curates a few thousand characters of notes about you. A brain gives it the other 99 percent, governed and recallable on demand.

Last updated

AIVM Brain connects to Hermes Agent as an MCP server, giving it governed recall over your full knowledge: projects, decisions, documents, history. Hermes ships a built-in MCP client, so the setup is one config entry. The brain works alongside Hermes's own memory files; it is independent infrastructure you own, not a Hermes component.

How do you connect Hermes Agent?

Connect Hermes Agent using the current setup instructions inside your AIVM Brain workspace. Public access currently starts with the waitlist.

  1. Get workspace access. Join the waitlist with your work email and confirm your email. If you already have access, log in to your workspace.

  2. Follow the in-app setup for Hermes Agent. Use the current connection instructions provided in your workspace for Hermes Agent. Follow the access scope assigned to your agent.

  3. Check access and retrieval. Ask Hermes Agent a question about an approved source and check its citation. Also test a restricted source to confirm that the agent cannot retrieve content outside its permissions.

Bounded files are a feature. They are also a ceiling.

Hermes Agent's built-in memory is deliberately small: as of mid-2026 it curates MEMORY.md and USER.md, bounded files injected into the system prompt at session start. That design keeps every session cheap and predictable, and for who-you-are context it works. But a few thousand characters cannot hold a year of decisions, a team's documents, or the reasoning behind last quarter's architecture change.

The brain is the deep store on the other side of that ceiling. Hermes keeps its quick notes; when a task needs the full picture, it queries the brain over MCP and pulls exactly the knowledge that clears its permissions. Small working memory, large governed recall. Same pattern your own head uses.

An always-on agent needs an accountable memory

Hermes is built to run continuously, respond on your channels, schedule its own work, and improve itself with skills. An agent with that much autonomy reading and writing your knowledge is exactly where governance stops being paperwork. With the brain, every recall and every capture Hermes makes is permission-checked and written to a tamper-evident, content-blind ledger. If Hermes did something with your data at 3am, you can see precisely what it touched.

Questions, answered

Is AIVM Brain one of Hermes's pluggable memory providers?

No. Hermes manages its own memory system separately. AIVM Brain connects over MCP as an external knowledge tool, alongside whatever memory setup Hermes runs.

Does this replace MEMORY.md and USER.md?

No, and it should not. Hermes's bounded files are its fast working notes. The brain is the deep, governed store for everything that does not fit in a system prompt.

Can multiple Hermes instances share one brain?

Yes. Each instance gets its own key, each key is scoped by role, and every access from every instance lands in the same ledger.

Is this self-hostable like Hermes?

The brain is a hosted service with per-tenant Postgres isolation and bring-your-own-model keys. Your Hermes stays wherever you run it; only the knowledge store is hosted.

What if Hermes changes its config format?

The brain speaks standard MCP, so any Hermes version with an MCP client can connect. Only the config file details are Hermes's; check its docs after upgrades.

Give Hermes Agent a brain it can be held to.