Card generation moved to response step:
- Response LLM outputs JSON with "text" + optional "card"
- Cards use actual query data, not placeholder templates
- Plan step no longer includes emit_card (avoids {{template}} syntax)
- Fallback: raw text response if JSON parse fails
History restore on reconnect:
- Frontend fetches /api/history on WS connect
- Renders last 20 messages in chat panel
- Only restores if chat is empty (fresh load)
Graph animation:
- Dynamic node name → graph ID mapping from graph definition
- All nodes (including eras_expert) pulse correctly
- 200ms animation queue prevents bulk event overlap
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Eras Expert domain context:
- Full Heizkostenabrechnung business model (Kunde>Objekte>Nutzeinheiten>Geraete)
- Known PK/FK mappings: kunden.Kundennummer, objekte.KundenID, etc.
- Correct JOIN example in SCHEMA prompt
- PA knows domain hierarchy for better job formulation
Iterative plan-execute in ExpertNode:
- DESCRIBE queries execute first, results injected into re-plan
- Re-plan uses actual column names from DESCRIBE
- Eliminates "Unknown column" errors on first query
Frontend:
- Node inspector: per-node cards with model, tokens, progress, last event
- Graph switcher buttons in top bar
- Clear button in top bar
- Nodes panel 300px wide
- WS reconnect on 1006 (deploy) without showing login
- Model info emitted on context HUD events
Domain context test: 21/21 (hierarchy, JOINs, FK, PA job quality)
Default graph: v4-eras
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
