Web GUI Quickstart
Configure a model, select a workspace and finish your first task in the Web UI
This guide starts from an already running server (see installation). The dsh process uses its launching directory as the default filesystem location, but the new Web UI selects no workspace until you add one.
1. Configure a model
Open Settings → Models, enter your DeepSeek API key and save. The model routing takes effect immediately — no server restart needed.
Other providers and custom OpenAI-compatible endpoints are covered in configure models.
2. Select a workspace
Click Select Workspace, add the project directory where you launched dsh, then select it.
The session input stays disabled until a workspace is selected.
3. Run your first task
Start a session and send:
Summarize this repository and identify its main packages.
The agent can read and edit workspace files, run commands, delegate work and maintain plans. When an action needs approval under the current permission policy, the Web UI asks you first — review it before approving.
The flow at a glance
4. Keep going
- Configure models: DeepSeek, other providers and custom endpoints
- CLI & profiles: more launch modes
- Vibe Coding 101: the first lesson for newcomers
- FAQ: common questions at a glance