DP Code stores server-authoritative settings inDocumentation Index
Fetch the complete documentation index at: https://mintlify.com/Emanuele-web04/dpcode/llms.txt
Use this file to discover all available pages before exploring further.
~/.dpcode/userdata/settings.json. These settings control server-side behavior such as assistant streaming, thread environments, and provider configuration. You can edit the file directly or use the in-app Settings UI — both methods persist to the same file.
If
settings.json contains invalid JSON or fails schema validation, DP Code logs a warning and falls back to all defaults. Check your server logs if a setting does not appear to take effect.File location
Top-level fields
enableAssistantStreaming
Type: boolean — Default: false
Controls whether assistant responses are streamed token-by-token to the browser as they arrive. When set to false, the full response is buffered server-side before being forwarded to the client.
defaultThreadEnvMode
Type: "local" | "worktree" — Default: "local"
Sets the default environment mode for new threads. Use "local" to run agents against the working tree of the project directly. Use "worktree" to spin up a Git worktree per thread, keeping each agent session isolated.
addProjectBaseDirectory
Type: string — Default: ""
A path prefix that DP Code prepends when resolving project directories. Leave empty to use project paths as-is. Useful when all your projects share a common root (for example /home/you/work).
textGenerationModelSelection
Type: { provider, model } — Default: { "provider": "codex", "model": "gpt-5.4-mini" }
The model used for short text generation tasks such as producing Git commit messages. This is separate from the model used for interactive coding sessions.
| Field | Description |
|---|---|
provider | Provider key ("codex", "claudeAgent", "gemini", "grok", "kilo", "opencode", "cursor", "pi") |
model | Model slug understood by the selected provider |
codex → claudeAgent → gemini → kilo → opencode.
providers
Type: object
Provider-specific configuration such as binary paths, launch arguments, and custom model lists. Each key corresponds to a supported provider. See the Providers configuration page for the full reference.