- agentic chat loop (max 6 round-trips, budget re-checked between iterations) over 10 read-only, permission-delegated tools in src/services/ai-tools.ts - persist assistant tool trace in ai_chat_messages.content_json (+ migration) - consulted-tools chips in OdinThread; header now shows month spend only (budget cap hidden from the UI, server-side 402 guard unchanged) - seed: ai.use permission; Settings exposes the ai permission module - docs: REVIEW consolidation; deployment-guide.md superseded by docs/release.md Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
8 lines
365 B
SQL
8 lines
365 B
SQL
-- Odin Phase 2a: assistant turns carry structured metadata (tool-call trace,
|
|
-- later full content blocks). `content` stays the plain-text display
|
|
-- projection; `content_json` holds the structured form. Anticipated by the
|
|
-- Phase-2 design notes (2026-06-08).
|
|
|
|
-- AlterTable
|
|
ALTER TABLE `ai_chat_messages` ADD COLUMN `content_json` LONGTEXT NULL AFTER `content`;
|