- Track added model names in state
- Disable already-added models with 'Added' badge
- Clear selection after successful add
- Add Next button at bottom when models have been added
- Provider form submit button changed to 'Save & Next'
- Remove extra Next button from provider form
- Add search input to filter scanned models
- Add manual model name input for APIs that don't support scanning
- Add divider between scanned list and manual add section
- Add i18n translations for all locales
When navigating back to the provider form after already creating a
provider, show a Next button that skips to the model scan or config
layer without requiring re-submission.
Track maxReachedLayer so clicking Next from an earlier layer jumps
directly to the deepest previously visited layer, preserving state.
Reset tracking when the user changes their AI engine mode choice.
- Add initialValues and onValuesChange props to ProviderForm
- Save provider form data and selected models in wizard state
- Fix canProceed missing modelsAdded dependency
- Restore form data and selections when navigating back
Use a dedicated step2Layer state for sub-layer navigation within
Step 2, instead of resetting data flags. This preserves provider
form data, scan results, and model selections when going back.
After creating a custom provider, scan for available LLM models
and let the user select which ones to add before proceeding.
Support back navigation from config form to model scan to provider form.
Pass provider UUID from ProviderForm callback.
After creating a custom provider, automatically select local-agent as
the runner so the config form (with model selector) is displayed.
Require runner selection for canProceed in LLM+custom path.
Restructure the wizard's Step 2 (AI Engine) into a two-layer choice:
- Orchestrated Agent apps (Dify, n8n, Coze, etc.)
- Direct LLM usage (Space service or custom provider)
For direct LLM mode, users choose between LangBot Space (OAuth)
or adding their own model provider via a dedicated ProviderForm page.
Support intra-step back navigation across all layers.
Add Vite dev server proxy config for API requests.
Add i18n translations for all 8 locales.
The local-agent runner's system prompt is a static string with no
template-variable support, so the model had no anchor for "today" and
resolved relative time references (e.g. "this quarter", "latest")
against whichever period was best represented in training data instead
of the real date, sometimes confidently answering with stale
information for time-sensitive questions.
PreProcessor now appends a short, deterministically-computed
"Current date: ..." note to the system prompt on every request for
local-agent pipelines, alongside guidance to verify time-sensitive
facts with a search tool rather than answering from memory. The
existing skill-awareness prompt injection is refactored to share the
same append-to-system-prompt helper.
* feat: report independent instance and workspace identities
* test: include workspace in OAuth callback fixture
* ci: pin production cloud adapter to Space release
* fix: preserve authenticated Workspace telemetry attribution
* ci: pin production cloud adapter to final Space release
---------
Co-authored-by: dadachann <185672915+dadachann@users.noreply.github.com>
* fix(cloud): show owner model balance and enforce single owner
* fix(migrations): create owner index idempotently
---------
Co-authored-by: dadachann <185672915+dadachann@users.noreply.github.com>
langbot-dev