Ω omega

omega is a terminal-based AI assistant that reads files, runs commands, and edits code. It talks to LLM providers (Ollama, OpenAI, Anthropic) and streams responses in real time. Single binary, full- screen TUI, persistent session tree, context compaction, and a in-process module system.

The same binary also serves a VS Code-style web interface: a dock with draggable panes, live-streaming chat with retry and image attachments, sessions with search and pinning, settings, cron, logs, skills and a kanban board — all embedded, no separate frontend to run (omega serve). Conversations branch: every retry is a sibling variant you can switch between. Agents can ask mid-run via verdict cards (single- or multi-choice, or free text), finished tasks ring a bell, chats can be ephemeral (in-memory only, nothing on disk), and multiple named agents run side by side.

Warning: The shell tool executes commands the LLM generates with no sandboxing. Use Ollama for sensitive work — cloud providers receive file contents and command output.

Community

Join the Discord for release announcements and discussion.

Quick Start

git clone https://github.com/EndoTheDev/omega.git
cd omega
./build.sh        # or build.bat on Windows
bin/omega         # config.yaml.example lands in bin/ on first run;
                  # create bin/config.yaml for non-default settings

Or install via go install:

go install github.com/EndoTheDev/omega/cmd/omega@latest

go install puts the binary in ~/go/bin (or $GOBIN) — add it to PATH if it isn't already:

export PATH="$HOME/go/bin:$PATH"   # add to ~/.bashrc or ~/.zshrc

Releases also ship prebuilt binaries for all platforms (with the webui embedded) at github.com/EndoTheDev/omega/releases.

Runtime layout & update

omega's home is the executable's directory (override with OMEGA_HOME). Installed into a bin directory (go install, ~/.local/bin), the home moves to ~/.omega — config, example configs, the session store and logs live there. Update an installed copy with:

omega update

Architecture

HTTP channel -> agent (loop + tools) -> ai (provider streaming)

Layer	Responsibility
Agent	Multi-turn loop, tools, compaction, modules
Provider	Ollama + OpenAI + Anthropic streaming
CLI	Entry point, config loading, trust gate, Frontend dispatch

See docs/architecture.md for details.

Modules

All modules are in-process Go packages under modules/. See modules/README.md for the full list.

Documentation

Topic	File
Configuration	docs/configuration.md
Extensions	docs/modules.md
TUI Commands	docs/tui.md
Project Trust	docs/project-trust.md
Architecture	docs/architecture.md
Landing pipeline	docs/pipeline.md
Roadmap	docs/roadmap.md
Troubleshooting	docs/troubleshooting.md

License

MIT