Control surface
A spatial canvas displays agents, screens, branches, resources, and their connections.
Wega Labs · October
The operating system for AI collaboration.
October coordinates humans and AI agents in one workspace. It exposes shared context, task ownership, parallel execution, and live output.
CategoryAI collaboration OS
The problem / 01
AI agents run in separate interfaces.
Agents run in separate terminals, chats, IDE panels, and vendor applications. Shared context, task ownership, messaging, and status are inconsistent.
October provides one coordination layer for humans and agents.
Why an operating system / 02
Three system layers.
Coordination layer
The october-bus provides peer discovery, messaging, canvas state, task ownership, and status.
Execution layer
Local agent processes edit repositories in isolated worktrees connected to dev servers and previews.
October does not implement the models. It coordinates agent processes across providers, editors, local machines, and remote systems.
Native collaboration / 03
The october-bus
An MCP protocol that gives agents access to peers, messages, canvas state, task ownership, and execution status.
HumanIntent + direction
→OrchestratorResolve names, screens, ownership
→Screen A
Backend
Tests
Shared bus
discover peers · inspect canvas · message agents · report status · preserve provenance
System status / 04
What exists. What remains.
Built + shipping
Current desktop- Spatial infinite canvas✓
- Route detection + live previews✓
- Local coding-agent harnesses✓
- Integrated terminals✓
- Git worktree isolation✓
- Variant previews + PR flows✓
- Otto voice control✓
- Cross-platform CI gate✓
Agent roster
Local harnesses01Claude Code02Codex03Cursor04Grok05Gemini06opencode07Hermes08Cline09Pi
All listed agents can run locally. Their october-bus integrations are not equivalent.
In development
Named agent control
Direct a specific agent by name or by its assigned screen.
Protocol parity
Claude Code currently has the deepest bus integration. Other agents need the same discovery, messaging, and status functions.
Concurrent editing
Add explicit ownership, presence, provenance, and synchronization for simultaneous work.
Local and remote agents
Expose local and remote agents through the same control and coordination interfaces.
Human control / 05
Observe, direct, and verify agent work.
System state
Detected routes become live screens. Connections record context and responsibility.
Execution
Agents edit local repositories. Hot module replacement updates previews after file changes.
Isolation
Git worktrees separate concurrent agent sessions before changes are merged.
Provenance
The system records which agent owns a task, which files changed, and where output was produced.
The collaboration loop / 06
1Detect
→October identifies repositories, routes, frameworks, and running services.
2Assign
→Name agents, connect them to screens, and assign bounded tasks.
3Coordinate
→Agents exchange context and status through the bus while working in separate worktrees.
4Verify
Humans inspect previews, diffs, terminals, and task provenance.
A place, not a panel / 07
Spatial layouts provide stable references.
A persistent layout lets humans and agents refer to the same screen, position, and local context across tasks.
October by Wega Labs
One workspace for humans and AI agents.
Shared context, task ownership, execution, and verification.
Visit october.dev ↗