Telegram
Ship a conversational agent to a group or DM in a single deploy. Threads, commands and inline replies, handled by the managed runtime.
Open Source Core · Managed by Elux Space
The agent that leaves orbit
SpaceClaw is managed hosting for OpenClaw, the open-source AI-agent framework. Launch a Hangar, deploy an agent to Telegram, Discord or Slack, and let us run the runtime. No VPS. No SSH. No hardening — just your agents, in orbit.
$ curl -fsSL https://spaceclaw.elux.space/install.sh | bash
SpaceClaw Workspace — Beta Preview
Live runtime · 00:14
One agent · every surface
Any channel
Deploy once. Reach your people wherever they already talk — one memory follows across all of them.
Discord
Wire your OpenClaw agent into a server as a first-class bot. Slash commands, roles and per-channel context out of the box.
Slack
Bring the same agent into the workspace where the work happens. Also on WhatsApp and CLI — same brain, any front door.
Channels are one of the four config pillars — Runtime · Model · Skills · Channels — set on every agent you deploy.
#1
Launch a Hangar
A Hangar is your hosting pool. Pick a plan and it sets your agent slots, disk allowance, monthly task-minutes and the reset date — provisioned in one click, no server to rent or patch.
Everything you deploy lives inside the Hangar, metered transparently and billed monthly in USD from your Elux Space balance.
Agent slots
Disk
Task-minutes
Reset date
#2
Deploy in minutes
Choose your OpenClaw runtime, bring your model and provider, attach skills, pick a channel. That is the whole deploy — the four pillars, one form, live agent.
No SSH, no Dockerfiles, no hardening checklist. The managed runtime boots, connects and starts answering.
Runtime
Model
Skills
Channels
#3
Every channel
Telegram, Discord, Slack, WhatsApp and CLI — the same agent speaks on all of them. Configure once; SpaceClaw handles the adapters, webhooks and auth for each surface.
Add or swap a channel later without redeploying the brain behind it.
Telegram
Discord
Slack
WhatsApp
CLI
#4
Tasks multiplied
Runtime is metered in task-minutes. Every plan sets a per-task ceiling, a generous monthly quota and a grace buffer on top — so a busy week never means a hard stop.
Need more headroom? Attach a task-minute add-on any time; it’s billed in USD to the same Elux Space balance.
Per-task limit
Monthly quota
Grace buffer
Add-on
#5
One memory
A conversation started on Telegram continues on Slack. Your agent keeps one shared memory across every channel, so context, history and preferences never fracture between surfaces.
Open-source core, your data, your agent — SpaceClaw only runs it.
Shared context
Cross-channel
Your data
#6
Isolated sandboxing
Each agent runs in its own isolated sandbox inside your Hangar. Skills execute in a contained runtime with metered resources, so one agent can never reach into another’s state or spend.
Restart a runtime, roll a config, or tear an agent down — the blast radius stops at its own slot.
Contained runtime
Per-agent slots
Safe restarts
Billed monthly in USD
Pick a Hangar
Four tiers. Agents, disk and task-minutes scale together — auto-renewed each cycle.
| Plan | Solo | Duo | Squadron | Fleet |
|---|---|---|---|---|
| Price / mo | $9 | $16 | $24 | $39 |
| Agents | 1 agent | 2 agents | 3 agents | 4 agents |
| Hangar disk | 15 GB | 30 GB | 45 GB | 75 GB |
| Task limit | 15 min / task | 30 min / task | 45 min / task | 60 min / task |
| Monthly quota | 5,000 task-min +1,000 grace |
20,000 task-min +4,000 grace |
50,000 task-min +10,000 grace |
120,000 task-min +24,000 grace |
| Solo | Duo | Squadron | Fleet |
Auto-renew — each cycle renews automatically, billed in USD to your Elux Space balance. If a payment can’t clear, your Hangar enters a grace hold instead of stopping cold.
Task-minute add-on — top up any plan with extra task-minutes on demand, priced in USD; add-ons never expire mid-cycle.
Open-source core — OpenClaw stays open. SpaceClaw is the managed runtime around it — cancel and self-host any time.