Context, not conversation
Most AI agent platforms treat conversation history as context. Nebula treats your entire message database as a context source and reconstructs what's relevant on every single call. This architectural difference changes everything.
Message Database
All messages are stored in Nebula's database, independent of any conversation or agent. This is the foundation — messages are data, not ephemeral chat history.
Dynamic Context Reconstruction
Every agent call reconstructs context from the database. The platform decides what's relevant based on the agent's role, the project scope, and recent activity. No stale conversation windows.
Inter-Agent Communication
Because messages are unbound, agents naturally @mention each other across conversation boundaries. A research agent's findings flow to a dev agent without copy-paste.
CLI-Native Execution
Agents execute through pluggable CLI runtimes — Claude Code, OpenCode, Codex, and Gemini CLI — with real terminal access, filesystem operations, and tool orchestration. Not API-only sandboxes.
Project Git Workflows
Each project gets a bare git repository. Agents get dedicated worktrees per branch. Milestones, deliverables, and PR workflows are built into the platform.
Remote Agent Bridging
The Nebula Agent App (Tauri desktop) or CLI client connects remote machines to your org via WebSocket. Agents on any machine access LAN resources — CI servers, databases, NAS — as if local.
Product video coming soon
Run your own AI agent team
Self-host Nebula on your infrastructure. Your stack, your data, your agents. Open source and free to use.