AI that thinks alongside you, not just when you ask.
- Parallel agents — multiple agents work concurrently on the same task; you never wait for one to finish before the next starts
- Persistent code intelligence — BM25 + vector index built on first run; agents know your codebase, not just your last message
- Impact analysis — before touching a file, agents see what depends on it
- Long-term memory — decisions made last week are available this session (via MuninnDB)
- Live delegation — sub-agents stream their work into a thread panel while you stay in your main conversation
Multi-agent team chat — parallel agents delegating and streaming work in real time
Agent configuration — custom personas, tool access, connections, and memory
Connections — GitHub, Slack, Jira, AWS, and 20+ integrations
Skills Marketplace — browse and install community skills
Workflows — cron-scheduled automation pipelines
Models — manage local Ollama models or connect cloud providers
Requirements: Either Ollama (local models) or a cloud API key (Anthropic, OpenAI, etc.).
macOS (Homebrew):
brew install scrypster/tap/huginnmacOS / Linux (one-liner):
curl -fsSL https://huginn.sh/install.sh | shWindows (PowerShell):
irm https://huginn.sh/install.ps1 | iexOr grab a binary directly from GitHub Releases for your platform.
Build from source
Requires Go 1.25+.
git clone https://github.com/scrypster/huginn
cd huginn
go build -tags embed_frontend -o huginn .
sudo mv huginn /usr/local/bin/huginnPull default local models (Ollama):
ollama pull qwen3-coder:30b # planner — Chris
ollama pull qwen2.5-coder:14b # coder — Steve
ollama pull deepseek-r1:14b # reasoner — Mark# 1. Launch TUI in any project directory
huginn
# 2. Single-turn non-interactive query
huginn --print "what does the auth middleware do?"
# 3. Launch the web UI (opens at http://localhost:8421)
huginn tray
# 4. Run a specific named agent
huginn --agent Steve "refactor internal/db/query.go to use prepared statements"| Feature | Docs |
|---|---|
| Web UI (all views, thread panel, sessions, inbox) | docs/features/web-ui.md |
| Multi-agent (parallel delegation, thread panel, swarm) | docs/features/multi-agent.md |
| Custom agents (create agents with any persona, model, and tools) | docs/features/custom-agents.md |
| Spaces (DMs, Channels, team rooms, workstreams) | docs/features/spaces.md |
| Artifacts (structured outputs, review & download) | docs/features/artifacts.md |
| Sessions & persistent history | docs/features/sessions.md |
| Memory (context notes and MuninnDB) | docs/features/memory.md |
| Notepad (inject standing context into sessions) | docs/features/notepad.md |
| Skills (teach agents new behaviors) | docs/features/skills.md |
| Skills Registry (community skills browser) | docs/features/skills-registry.md |
| Routines & automation | docs/features/routines.md |
| Workflows (chain steps into pipelines) | docs/features/workflows.md |
| Connections (GitHub, Jira, Slack, MCP) | docs/features/connections.md |
| Code Intelligence (semantic search, impact analysis) | docs/features/code-intelligence.md |
| Permissions & safety | docs/features/permissions.md |
| HuginnCloud (access agents from anywhere) | docs/features/huginncloud.md |
| Headless / server mode | docs/features/headless.md |
| Slash commands | docs/features/slash-commands.md |
| CLI reference | docs/reference/cli.md |
| Config reference | docs/reference/config.md |
| TUI keybindings & slash commands | docs/reference/tui.md |
| Glossary | docs/reference/glossary.md |
| Troubleshooting | docs/troubleshooting.md |
Full docs index: docs/index.md
Huginn is source-available under the Business Source License 1.1.
- Free for: personal use, internal tooling, research, and non-commercial projects
- Commercial license required for: offering Huginn as a hosted service or embedding it in a commercial product
- Converts to Apache 2.0 on March 7, 2030
See docs/CONTRIBUTING.md. Build, test, and PR guidelines are all there.





