Hermes,
the messenger in your team.
OpenCognit runs Hermes Agent in real terminals or as a thread with approvals — next to Claude Code, Codex, Gemini and Kimi. Hermes can conduct the others, and every command it runs is checked first.
Hermes and the others,
in one window.
Hermes as you know it,
with a control center around it.
Terminal or thread
Run Hermes in its real terminal — or as a thread: a chat with approval cards for every action.
Hermes as the conductor
The orchestrator can run on Hermes: it starts agents, hands out tasks and reports back when they are done.
Checked before it runs
Hermes starts with a config folder of its own that links to your real one — only the config file gets the check hook. Your skills, sessions and history stay where they are.
Reviews next to the builders
Let Hermes review a branch while Claude builds and Kimi tests — each in its own terminal, all in one grid.
From agent to outcome
The Layers view shows which agent works in which area and what came out of it: files to review, tests, commits.
Speak instead of type
Hold Ctrl+Space, speak, let go — your words land in the Hermes terminal or chat.
Three steps, no new account.
Install Hermes
As usual, with your own setup.
Install OpenCognit
Linux (AppImage, .deb) or Windows. It finds Hermes by itself.
Open a project
Start Hermes as a terminal or a thread — or let it conduct.
Questions about Hermes
Does OpenCognit change my Hermes configuration?
No. Hermes starts with its own config folder that links to your real one; only the copy of the config file carries the check hook.
Can Hermes be the orchestrator?
Yes. Claude Code, Kimi Code and Hermes can conduct; you pick the CLI for each new chat.
Does the Usage page count Hermes?
Not yet — Hermes keeps no readable usage log.
Does it run on macOS?
Not yet. OpenCognit is available for Linux and Windows.
What does it cost?
Free with up to 4 agents and 2 projects. Pro is a one-time purchase — no subscription on top of your plan.
