mirror of
https://github.com/hellodigua/ChatLab.git
synced 2026-10-26 02:04:37 +08:00
diguaandCursor
3ab162162b
Extract LLM config, assistant, skill, conversation and summary CRUD
routes from CLI-specific ai.ts into @openchatlab/http-routes, enabling
Electron Internal Server to serve AI endpoints via shared routes.
Frontend service layer:
- Create src/services/{llm,assistant,skill}/ with typed adapters
- Migrate all window.llmApi/assistantApi/skillApi calls to service layer
- Clean up Electron preload: remove CRUD bridges, keep only streaming IPC
Shared AI routes (packages/http-routes):
- ai-llm: config CRUD, provider registry, model catalog, custom store
- ai-assistants: CRUD, import/reimport, tool catalog
- ai-skills: CRUD, import/reimport, builtin catalog
- ai-conversations: conversation/message/debug CRUD
- ai-summaries: summary generation with LLM config injection
- Add requireAi fail-fast option to registerSharedRoutes
Backend improvements:
- Extract CustomProviderStore/CustomModelStore to node-runtime
- Fix authProfile persistence (onApiKeyCreated returns profileName)
- Reuse AI manager singletons in Electron (avoid dual instances)
- Prevent plaintext API key leakage in addConfig response
- Add authProfile regression tests
Co-authored-by: Cursor <cursoragent@cursor.com>