Frequently Asked Questions
Find answers to common questions about Stib. If your question isn't covered here, check the Troubleshooting page or contact us.
General
What is Stib?
Stib is a Claude Code agent orchestrator with a visual Kanban board. It lets you spawn, monitor, and interact with Claude Code agents through an intuitive interface — managing AI-driven development tasks the same way you manage any project board.
Stib runs as a self-hosted server and is available as a Docker image, pre-compiled binary, or desktop app. A companion mobile app lets you manage your agents on the go.
What is Claude Code?
Claude Code is Anthropic's agentic coding tool. It runs in a terminal and can read, write, and execute code autonomously. Stib wraps Claude Code with a visual interface, persistent sessions, and pipeline automation.
What platforms are supported?
Stib runs on macOS, Linux, and Windows. The desktop app includes system tray integration for all three platforms. The mobile app runs on iOS and Android.
Is Stib open source?
Stib is a commercial product with a proprietary license. Check the website for the full license terms.
What are the system requirements?
Stib requires:
- Claude Code CLI installed and authenticated
- Git (optional, for worktree features)
- Minimal CPU and RAM — Stib itself is lightweight; Claude Code agents do the heavy lifting via API
Installation
Which installation method should I choose?
| Method | Best for |
|---|---|
| Docker | Quick setup, isolated environment, server deployments |
| Docker Compose | Production setups with custom configuration |
| Binary | Direct installation without containers |
| Desktop | Personal use with system tray integration |
What is the default port?
Stib listens on port 50505 by default. The port is fixed and not configurable.
Where is my data stored?
All data is stored in a SQLite database in Stib's data directory. The default location depends on your installation method:
- Docker: Inside the container at
/data(mount a volume to persist) - Binary/Desktop: In the OS-specific data directory (e.g.,
~/.local/share/stibon Linux,~/Library/Application Support/stibon macOS)
TIP
When using Docker, always mount a volume (-v stib-data:/app/data) to preserve your data across container restarts.
→ Configuration and data paths
How do I update Stib?
The update process depends on your installation method:
- Docker: Pull the latest image and restart the container
- Binary: Download the new release and replace the binary
- Desktop: Download the latest version from the download page and install it over the existing version
Can I use a reverse proxy?
Yes. Stib works behind reverse proxies like Nginx, Caddy, or Traefik. Make sure to:
- Forward WebSocket connections (used for real-time sync)
- Set the
STIB_SERVER_ORIGINenvironment variable to your public URL - Pass through the
UpgradeandConnectionheaders for WebSocket support
WARNING
WebSocket forwarding is required. Without it, real-time updates (agent output, card moves) will not work.
Agents & Pipeline
What happens if an agent fails?
If a Claude Code agent encounters an error or crashes, the card remains in its current column and the error state is displayed on the card. You can retry the agent from the card's action menu. Session history is preserved so the agent can resume with full context.
What is context saturation?
Context saturation occurs when a Claude Code agent's context window fills up. When this happens, Stib automatically performs a context relay: it spawns a fresh agent and transfers the conversation context so work continues seamlessly. No manual intervention is needed.
How does auto mode work?
Auto mode lets you define an automatic pipeline progression. When an agent completes its work in one column, Stib automatically moves the card to the next column and spawns a new agent. You can configure which columns are chained and set stop points where the pipeline pauses for human review.
Can I cancel a running agent?
Yes. You can cancel a running agent at any time from the card's overlay panel or from the card's action menu. The agent process is terminated and the card stays in its current column. You can restart the agent afterwards.
Do agents survive page refresh?
Yes. Agent sessions are fully persistent. If you close the browser, refresh the page, or disconnect, the agent continues running on the server. When you reconnect, you see the full conversation history and can resume interaction immediately.
Git & Worktrees
What is a worktree?
A Git worktree is a separate working directory linked to the same repository. Stib creates a dedicated worktree for each card, giving every agent its own isolated copy of the code. This prevents agents from interfering with each other's changes.
How are branches named?
Stib creates branches following the pattern stib/<card-id>-<card-title-slug>. Each card gets a unique branch in its own worktree, keeping changes organized and easy to identify.
When are worktrees cleaned up?
Worktrees are automatically cleaned up when a card reaches the final column (typically "Done"). You can also manually clean up worktrees from the card's action menu. The associated branch is preserved unless you explicitly delete it.
Can I use Stib without Git?
Yes. Git integration is entirely optional. If your project directory is not a Git repository, Stib simply skips worktree creation and branch management. All other features — agents, pipeline, Kanban board — work normally.
Mobile
What devices are supported?
The Stib mobile app runs on iOS and Android. It requires a network connection to your Stib server.
How do I connect the mobile app?
You can connect in two ways:
- QR Code — Open Settings in the Stib web UI and scan the QR code with the mobile app
- Manual entry — Enter your server URL and authentication credentials directly in the app
What can I do on mobile?
The mobile app lets you:
- Browse projects and boards
- View the Kanban board and card details
- Create cards and move them between columns
- Chat with running agents
- Monitor agent execution in real-time
Does the mobile app work offline?
No. The mobile app requires a network connection to your Stib server. All data is stored on the server and the app streams updates in real-time via WebSocket.
Account & Authentication
How do I set up SSO?
Stib supports OIDC Single Sign-On with any compliant identity provider. Configuration is done from the Settings page in the web UI. Stib includes presets for common providers like Google, Microsoft Entra ID, and Keycloak.
Can multiple users share a Stib instance?
Yes. When authentication is enabled, multiple users can log in and collaborate. Projects can be shared with specific users or scoped to organizations. Each user sees only the projects they have access to.
What OIDC providers are supported?
Stib works with any OIDC-compliant provider, including:
- Google (Google Workspace)
- Microsoft Entra ID (Azure AD)
- Keycloak
- Auth0
- Okta
- Any provider that implements the OpenID Connect standard
How do I manage backups?
Stib includes a built-in backup feature accessible from Settings. You can create manual backups or configure periodic automatic backups. Backups include the SQLite database and configuration data.
Next Steps
Can't find what you're looking for? Check the Troubleshooting & Logs guide for diagnosing common issues, or browse the full documentation for in-depth coverage of every feature.