- New AgentPipeline model with JSONB pipeline_data for stages/leads/summary
- AgentOrchestrator service chains DiscoveryService search→analyze→outreach→auto-save
- 3 new API endpoints: POST /agent/start, GET /agent/pipelines, GET /agent/{id}
- Full Agent dashboard Vue component with stats, pipeline grid, leads table, outreach preview
- Sidebar redesigned with AI Agent as primary entry point
- Updated PROGRESS.md, AGENTS.md, DATABASE_SCHEMA.md with latest state
- Make AI routing rules DB-driven (read from system_configs, removed from config.py)
- Add translation quota tracking to LLM translation (OpenAIProvider)
- Add Alibaba MT ECS RAM role support (STS token, no AccessKey needed)
- Fix admin sidebar link for AI模型配置 page
- Fix Quota.vue API path (quotas → translation-quotas)
- Fix login auto-redirect to dashboard
- Add provider dropdown selects to AI routing config UI
- Clean up stale ai_provider_* system_configs records
- Remove OpencodeGo, Spark providers (code + DB)
- Update deploy config: nginx port 8000, systemd cwd
- Save every search result to DB for later review
- Add '搜索历史' tab with timeline view, load/delete records
- Raise discovery search timeout from 30s to 120s (Bing Puppeteer needs ~40s)
- Reduce search queries from 4 to 3 for faster response
- New model: DiscoveryRecord (user_id, product, market, companies JSON)
- API: POST/GET/DELETE /api/v1/discovery/records
- Migration: discovery_records table
- Separate workspace landing from login for better UX
- Referral system rewards both parties with Pro days
- Quota enforcement prevents abuse without breaking endpoints
- 7-day free trial with auto-downgrade on expiry
- Admin-managed search provider config (SearXNG, Bing)
- 15% discount on annual subscriptions
- MCP search server wrapping opencode search
- Fix discovery module field name mismatch causing 422
wlt
TradeMate Dev