ai: per-config reasoning_effort + composer THINK chip

Two layers for thinking-mode control:

1. Per-config default (Settings → LLM)
   New "Reasoning effort" Select in the Add/Edit dialog with
   off/low/medium/high/max + a budget hint per option (~2k, ~8k,
   ~24k, ~64k thinking tokens). Saved row meta line surfaces the
   level inline so it's visible without opening the editor.

2. Per-message override (composer chip)
   New ReasoningChip next to the model picker. Click cycles through
   the same five levels. Hidden chrome when off (muted "think" pill);
   sodium-amber active style with the level label when set.

   Persisted to crema.ai.reasoning so a refresh keeps the operator's
   intent, wiped together with the conversation on Clear.

When sending, withReasoning() merges reasoning_effort into the request
body as a top-level field. The proxy forwards it untouched to OpenAI /
DeepSeek (native field) and translates to Anthropic's thinking block
server-side.

reasoningEffortRef sidesteps a useCallback ordering issue —
regenerateLast/continueLast are declared before the state hook, so
they read the ref instead of a stale closure.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

app/lib/arcadia/llm-configs.ts

@@ -12,6 +12,20 @@ import type { ArcadiaClient } from "@crema/arcadia-client"
 
 export type LlmProvider = "openai" | "anthropic" | "deepseek" | "qwen" | "lmstudio"
 
+/**
+ * Reasoning effort. Sent verbatim to OpenAI / DeepSeek (which take
+ * `reasoning_effort` natively). Translated server-side into Anthropic's
+ * thinking block. `off` (or null) skips the field entirely.
+ */
+export type ReasoningEffort = "off" | "low" | "medium" | "high" | "max"
+export const REASONING_EFFORTS: ReasoningEffort[] = [
+  "off",
+  "low",
+  "medium",
+  "high",
+  "max",
+]
+
 export interface LlmConfiguration {
   id: string
   tenant_id: string | null
@@ -23,6 +37,7 @@ export interface LlmConfiguration {
   input_cost_per_million: number | null
   output_cost_per_million: number | null
   enabled: boolean
+  reasoning_effort: ReasoningEffort | null
   metadata: Record<string, unknown>
   inserted_at: string
   updated_at: string
@@ -39,6 +54,7 @@ export interface LlmConfigurationInput {
   input_cost_per_million?: number | null
   output_cost_per_million?: number | null
   enabled?: boolean
+  reasoning_effort?: ReasoningEffort | null
   metadata?: Record<string, unknown>
 }