HivemindOS manual
Slash Commands
Type / in a HivemindOS chat to run a command. There are two kinds:
- Dashboard commands, handled by HivemindOS itself.
- Hermes and gateway commands, handled by the connected agent runtime.
Some runtime commands only work in that runtime’s own CLI or gateway, but may still appear in the dashboard’s help. When both kinds share a name, such as /sessions or /usage, the dashboard command runs in HivemindOS chat. Use the runtime’s own CLI for its version.
Dashboard commands
| Command | Aliases | Purpose |
|---|---|---|
/doctor |
/diagnostics |
Open fleet diagnostics and run checks. |
/sessions |
/history |
Search recent runtime sessions. |
/fleet |
Open the fleet dashboard. | |
/models |
/providers |
Open agent model/provider settings. |
/usage |
Open Wallets and refresh runtime usage. | |
/alerts |
/notifications |
Open the shared alert inbox. |
/brain |
/vault |
Open the shared brain. |
/note <note> |
Save a note to the shared brain intake folder. | |
/work |
/kanban |
Open the work board. |
/handoff-task <machine> <task> |
Send a task to the best agent on a target machine. | |
/swarm [number] <task> |
Have a team of your chat-capable agents work on the task in parallel and return one combined result. Leave out the number to size the team automatically. | |
/swarm-goal <build request> |
Turn a rough build request into a parallel /goal plan and send it to Queen Bee, which checks the result with browser, artifact, and independent review gates. |
|
/swarm-sim <scenario> |
Start a MiroShark simulation and show its queued status. | |
/image-gen <prompt> |
/imagine, /txt2img |
Generate an image and show it in chat. |
/video-gen <prompt> |
/txt2vid, /animate |
Generate a video and show it in chat. An attached picture becomes the first frame. |
Choosing where media is generated
When you pick /image-gen or /video-gen from the slash menu, a picker asks where to generate. If you type the prompt directly, it uses Automatic.
- This machine: free local options. Nothing leaves the machine.
- Other machines: free options on your other connected machines.
- HivemindOS credits: hosted models paid from your shared credit balance. Each row shows its cost in credits.
- Your accounts: models billed to accounts you already pay for, shared through Hivemind Content Studio.
Automatic prefers free and local, then credits, then your accounts, and tells you what it chose. Your last pick is remembered; if it isn’t available, HivemindOS falls back to Automatic and says so. Models that start from a picture need one attached to your message. Hosted picture-to-video models can’t run from chat yet.
Agents can read the same model list with the media_model_catalog tool.
AEON operator: /aeon
/aeon [request] turns the chat into an AEON operator session for your linked AEON workspace. Use it to set up an instance, edit skills or schedules, find out why a run was missed, turn the current chat into a scheduled automation, or edit STRATEGY.md, soul/SOUL.md, or soul/STYLE.md.
/aeon
/aeon move the digest to 7am Manila time
/aeon turn this chat into a weekday automation
/aeon tighten STRATEGY.md around qualified pipeline
/aeon make the writing voice less formal
- In a company CEO chat, it uses that company’s existing AEON workspace. It never creates a second workspace or changes the company’s link.
- It works the same when the workspace is on another linked machine.
- Creating repositories, adding credentials, real dispatches, commits, and pushes still ask for confirmation as usual.
Hermes and gateway commands
| Command | Aliases | Scope | Purpose |
|---|---|---|---|
/aeon [request] |
Autopilot | Set up or operate a linked AEON workspace; company-scoped chat uses the company’s AEON backbone. | |
/title [name] |
Session | Set a title for the current session. | |
/handoff <platform> |
CLI | Hand off this session to a messaging platform. | |
/branch [name] |
/fork |
Session | Branch the current session. |
/compress [focus topic] |
Session | Manually compress conversation context. | |
/rollback [number] |
Session | List or restore filesystem checkpoints. | |
/snapshot [create\|restore <id>\|prune] |
/snap |
CLI | Create or restore Hermes config/state snapshots. |
/stop |
Session | Kill running background processes. | |
/approve [session\|always] |
Gateway | Approve a pending dangerous command. | |
/deny |
Gateway | Deny a pending dangerous command. | |
/background <prompt> |
/bg, /btw |
Session | Run a prompt in the background. |
/agents |
/tasks |
Session | Show active agents and running tasks. |
/queue <prompt> |
/q |
Session | Queue a prompt for the next turn. |
/steer <prompt> |
Session | Inject a note after the next tool call. | |
/goal [text\|pause\|resume\|clear\|status] |
Session | Set or manage a persistent goal. | |
/subgoal [text\|remove N\|clear] |
Session | Add or manage criteria on the active goal. | |
/status |
Session | Show session info. | |
/whoami |
Info | Show slash command access. | |
/profile |
Info | Show active profile and home directory. | |
/sethome |
/set-home |
Gateway | Set this chat as the home channel. |
/resume [name] |
Session | Resume a named session. | |
/sessions |
Session | Browse and resume previous sessions in native Hermes surfaces. | |
/config |
CLI | Show current configuration. | |
/model [model] [--provider name] [--global] |
/provider |
Configuration | Switch model for this session. |
/codex-runtime [auto\|codex_app_server] |
/codex_runtime |
Configuration | Toggle Codex app-server runtime. |
/gquota |
CLI | Show Google Gemini Code Assist quota usage. | |
/personality [name] |
Configuration | Set a predefined personality. | |
/statusbar |
/sb |
CLI | Toggle the context/model status bar. |
/verbose |
CLI | Cycle tool progress display. | |
/footer [on\|off\|status] |
Configuration | Toggle gateway runtime metadata footer. | |
/yolo |
Configuration | Toggle approval-free YOLO mode. | |
/reasoning [level\|show\|hide] |
Configuration | Manage reasoning effort and display. | |
/fast [normal\|fast\|status] |
Configuration | Toggle provider fast mode. | |
/skin [name] |
CLI | Show or change the display skin/theme. | |
/indicator [kaomoji\|emoji\|unicode\|ascii] |
CLI | Pick the TUI busy indicator style. | |
/voice [on\|off\|tts\|status] |
Configuration | Toggle voice mode. | |
/busy [queue\|steer\|interrupt\|status] |
CLI | Control Enter behavior while Hermes is working. | |
/tools [list\|disable\|enable] [name...] |
CLI | Manage tools. | |
/toolsets |
CLI | List available toolsets. | |
/skills |
CLI | Search, install, inspect, or manage skills. | |
/bundles |
Tools & Skills | List skill bundles and aliases. | |
/cron [subcommand] |
CLI | Manage scheduled tasks. | |
/curator [status\|run\|pin\|archive] |
Tools & Skills | Background skill maintenance. | |
/kanban [subcommand] |
Tools & Skills | Use the collaboration board. | |
/reload |
CLI | Reload environment variables. | |
/reload-mcp |
/reload_mcp |
Tools & Skills | Reload MCP servers from config. |
/reload-skills |
/reload_skills |
Tools & Skills | Re-scan installed skills. |
/browser [connect\|disconnect\|status] |
CLI | Connect browser tools through CDP. | |
/plugins |
CLI | List installed plugins and status. | |
/commands [page] |
Gateway | Browse all commands and skills. | |
/help |
Info | Show available commands. | |
/restart |
Gateway | Gracefully restart the gateway. | |
/usage |
Info | Show token usage and rate limits in native Hermes surfaces. | |
/insights [days] |
Info | Show usage analytics. | |
/platforms |
/gateway |
CLI | Show gateway platform status. |
/platform <pause\|resume\|list> [name] |
Gateway | Pause, resume, or list a gateway platform. | |
/copy [number] |
CLI | Copy the last assistant response. | |
/paste |
CLI | Attach a clipboard image. | |
/image <path> |
CLI | Attach a local image file. | |
/update |
Info | Update Hermes Agent. | |
/debug |
Info | Upload a debug report. | |
/quit [--delete] |
/exit |
Exit | Exit the CLI. |
/<skill-name> |
Dynamic | Invoke an installed Hermes skill by name. |