refactor(core): upgrade ai-sdk v4 → v6 + Anthropic/openai-compatible reasoning round-trip + max-thinking budget audit

Migrates the LLM stack from [email protected] + @ai-sdk/[email protected] +
@ai-sdk/[email protected] to [email protected] + @ai-sdk/[email protected]
+ @ai-sdk/[email protected]. Full design in plan-v6-upgrade.md;
two rounds of Gemini code review captured in report.md.

Motivation: the recurring 'reasoning-signature without reasoning' error
on Claude Opus 4.7 was a v4 SDK artefact — @ai-sdk/[email protected] emitted
Anthropic signature_delta as a separate stream chunk that orphaned when
the model produced a signed-but-empty thinking block, and our chunk
store had no signature field so the round-trip back to Anthropic was
rejected on the next turn. In v6, signatures arrive inside
providerMetadata on the reasoning-end event, and the orphan-signature
class of bug is gone at the SDK level.

Core changes:
  • ThinkingChunk gains optional metadata?: Record<string, unknown>
    (the v6 providerMetadata blob). A non-undefined metadata 'seals'
    the chunk: subsequent reasoning-delta opens a new chunk rather
    than extending the sealed one.
  • AgentEvent gains { type: 'reasoning-end'; metadata? } (replaces
    the v4 reasoning-signature variant).
  • toModelMessages (replaces toCoreMessages):
      - returns ModelMessage[] (was CoreMessage[])
      - thinking → { type: 'reasoning', text, providerOptions: metadata }
      - tool-batch entries → { type: 'tool-call', input } (was 'args')
      - tool results → { output: { type: 'text', value } } ToolResultOutput
  • Claude OAuth uses createAnthropic({ authToken }) natively — no more
    custom-fetch x-api-key → Bearer swap.
  • rewriteBodyForOpus47 deleted — Opus 4.7 adaptive thinking is native
    via providerOptions.anthropic.thinking = { type: 'adaptive' }.
  • V1 middleware → V3 (specificationVersion: 'v3').
  • v4-era normalizeMessages openai-compatible middleware deleted; the
    v6 openai-compatible provider extracts reasoning_content natively
    from { type: 'reasoning' } content parts.
  • applyAnthropicStructuralNormalisations (mirrors opencode
    provider/transform.ts:53-148): drops empty text/reasoning parts,
    scrubs non-[a-zA-Z0-9_-] toolCallIds, splits [tool-call, non-tool]
    assistant turns (Anthropic rejects tool_use followed by text).
  • applyOpenAICompatibleReasoningNormalisation (mirrors opencode
    transform.ts:217-249): lifts reasoning text into
    providerOptions.openaiCompatible.reasoning_content (always, even
    empty). Solves DeepSeek 'The reasoning_content in the thinking
    mode must be passed back' — the v6 SDK skips emitting
    reasoning_content when text is empty (dist/index.mjs:245), but
    DeepSeek requires the field present once thinking was used.
  • Tools: tool({ inputSchema: jsonSchema(zodToJsonSchema(...)) })
    (was parameters: ZodSchema). AI SDK tools have no execute
    callback — the agent runs tools manually for permission prompts
    and shell-output streaming. New dep: zod-to-json-schema@^3.25.2.
  • fullStream event loop rewritten for v6 event shape: text-delta
    (text not textDelta), reasoning-start/delta/end, tool-input-*,
    tool-call (input not args), tool-result, tool-error (new), abort
    (new), start-step/finish-step, finish.

Max-thinking audit (matches opencode transform.ts:642-671 budgets):
  • Claude enabled-thinking max budget 16000 → 31999 (Anthropic ceiling)
  • Claude enabled-thinking high budget 10000 → 16000
  • maxOutputTokens 'budget + 8000' → fixed 32000 (matches opencode's
    OUTPUT_TOKEN_MAX; model self-allocates thinking vs response within)
  • Opus 4.7 adaptive thinking gains display: 'summarized' and sibling
    effort field (without these, thinking content is hidden by Anthropic
    and the model barely thinks).

Frontend mirrors:
  • types.ts — ThinkingChunk.metadata?, AgentEvent reasoning-end
  • tabs.svelte.ts — routes reasoning-end through applyChunkEvent
  • ChatMessage.svelte — hides empty thinking chunks; hides the entire
    assistant bubble when no chunk has renderable content

Gemini-review-driven fixes:
  • tool-error and abort stream events now surface as error chunks
    (were silently ignored)
  • toolCallId scrubbing pass (opencode transform.ts:96-122 parity)
  • Empty-reasoning-cull explicit test coverage for both Anthropic
    structural normalisation and DeepSeek path

Test counts (223 tests across 3 packages, all green):
  • tests/chunks/append.test.ts: 44 (was 38) — reasoning-end sealing,
    orphan walk-back, multi-block interleaving
  • tests/agent/agent.test.ts: 24 (was 5) — exhaustive v6 event
    mappings, structural normalisations, signature/reasoning_content
    round-trip, tool-error/abort branches, DeepSeek scenario, empty
    reasoning edge case
  • tests/llm/provider.test.ts: 9 (was 22) — dropped 13 obsolete v4
    middleware tests; new minimal tests confirm no middleware wrapping
    on default openai-compat path and that createAnthropic gets
    authToken vs apiKey correctly for OAuth vs api-key flows
  • tests/tools/registry.test.ts: 10 (was 4) — v6 tool() contract
    (inputSchema, no execute, JSON Schema for nested zod)
  • packages/api/tests/agent-manager.test.ts: 12 (was 7) — mock Agent
    emits v6 reasoning events; reasoning-end broadcast + ordering
  • packages/frontend/tests/chat-store.test.ts: 35 (was 32) —
    reasoning-end flow through Svelte $state store

typecheck clean (tsc --noEmit on core + api, svelte-check on frontend),
biome clean across 124 files.

3 files changed, 62 insertions, 6 deletions

diff --git a/packages/frontend/src/lib/components/ChatMessage.svelte b/packages/frontend/src/lib/components/ChatMessage.svelte
index 0c85349..54e99c8 100644
--- a/packages/frontend/src/lib/components/ChatMessage.svelte
+++ b/packages/frontend/src/lib/components/ChatMessage.svelte
@@ -36,6 +36,42 @@ const SYSTEM_KIND_LABEL: Record<SystemChunkKind, string> = {
 	"config-reload": "Config reload",
 	cancelled: "Cancelled",
 };
+
+/**
+ * Returns true if the given chunk has visible content worth rendering.
+ * Used by `hasRenderableContent` to suppress empty assistant bubbles.
+ *
+ * Note: `ThinkingChunk.metadata` is intentionally excluded — it is
+ * internal wire data (Anthropic's providerMetadata / signature) and
+ * must never appear in the UI.
+ */
+function chunkHasRenderableContent(chunk: Chunk): boolean {
+	switch (chunk.type) {
+		case "text":
+			return chunk.text.length > 0;
+		case "thinking":
+			return chunk.text.length > 0;
+		case "tool-batch":
+			return chunk.calls.length > 0;
+		case "error":
+			return true;
+		case "system":
+			return true;
+	}
+}
+
+/**
+ * True when the assistant bubble has something worth showing.
+ * Guards the assistant render path so we don't emit an empty box
+ * (e.g. a message that only had empty/signature-only thinking blocks
+ * from Anthropic adaptive thinking mode).
+ *
+ * Streaming messages always have renderable content — the cursor
+ * needs somewhere to live.
+ */
+const hasRenderableContent = $derived(
+	message.isStreaming === true || message.chunks.some(chunkHasRenderableContent),
+);
 </script>
 
 {#snippet renderChunks(chunks: Chunk[], streaming: boolean | undefined)}
@@ -43,13 +79,18 @@ const SYSTEM_KIND_LABEL: Record<SystemChunkKind, string> = {
 		{#if chunk.type === "text"}
 			<MarkdownRenderer text={chunk.text} {streaming} />
 		{:else if chunk.type === "thinking"}
-			<div class="collapse collapse-arrow mb-2 p-1">
-				<input type="checkbox" checked={appSettings.autoExpandThinking} />
-				<div class="collapse-title text-sm opacity-60 italic py-0 pl-0 pr-8 min-h-0">Thinking...</div>
-				<div class="collapse-content text-sm opacity-60 italic p-0">
-					<p class="whitespace-pre-wrap mt-1">{chunk.text}</p>
+			<!-- Skip empty thinking chunks: Anthropic adaptive thinking can emit
+			     a reasoning-end with a signature but no thinking_delta content.
+			     The metadata is internal wire data — never displayed. -->
+			{#if chunk.text.length > 0}
+				<div class="collapse collapse-arrow mb-2 p-1">
+					<input type="checkbox" checked={appSettings.autoExpandThinking} />
+					<div class="collapse-title text-sm opacity-60 italic py-0 pl-0 pr-8 min-h-0">Thinking...</div>
+					<div class="collapse-content text-sm opacity-60 italic p-0">
+						<p class="whitespace-pre-wrap mt-1">{chunk.text}</p>
+					</div>
 				</div>
-			</div>
+			{/if}
 		{:else if chunk.type === "tool-batch"}
 			{#each chunk.calls as call (call.id)}
 				<ToolCallDisplay toolCall={call} />
@@ -78,6 +119,11 @@ const SYSTEM_KIND_LABEL: Record<SystemChunkKind, string> = {
 			{@render renderChunks(message.chunks, false)}
 		</div>
 	</div>
+{:else if !isUser && !hasRenderableContent}
+	<!-- Empty assistant message — no renderable chunks and not streaming.
+	     Suppressed to avoid an empty bubble (e.g. a turn that produced
+	     only empty/signature-only thinking blocks from Anthropic adaptive
+	     thinking mode, or a done event with no content). -->
 {:else}
 <div class="chat chat-start mb-2 [&>.chat-bubble]:max-w-full {isQueued ? 'opacity-60' : ''}">
 	<div class="chat-bubble break-words {isUser ? 'chat-bubble-primary w-fit' : 'bg-transparent w-full'}">
diff --git a/packages/frontend/src/lib/tabs.svelte.ts b/packages/frontend/src/lib/tabs.svelte.ts
index b07d37a..1d60e0b 100644
--- a/packages/frontend/src/lib/tabs.svelte.ts
+++ b/packages/frontend/src/lib/tabs.svelte.ts
@@ -445,6 +445,7 @@ export function createTabStore() {
 				break;
 			}
 			case "reasoning-delta":
+			case "reasoning-end":
 			case "text-delta":
 			case "tool-call":
 			case "tool-result":
diff --git a/packages/frontend/src/lib/types.ts b/packages/frontend/src/lib/types.ts
index 1043f64..6051810 100644
--- a/packages/frontend/src/lib/types.ts
+++ b/packages/frontend/src/lib/types.ts
@@ -28,6 +28,14 @@ export interface TextChunk {
 export interface ThinkingChunk {
 	type: "thinking";
 	text: string;
+	/**
+	 * Mirror of core. Anthropic's `providerMetadata` blob captured from
+	 * the v6 `reasoning-end` stream event. Present once the backend has
+	 * sealed the chunk; absent for in-flight thinking or for non-Anthropic
+	 * models. The UI doesn't render this — it lives here for wire-format
+	 * symmetry with the persisted chunk shape.
+	 */
+	metadata?: Record<string, unknown>;
 }
 
 export interface ToolBatchChunk {
@@ -77,6 +85,7 @@ export type AgentEvent =
 	| { type: "statuses"; statuses: Record<string, "idle" | "running" | "error"> }
 	| { type: "text-delta"; delta: string }
 	| { type: "reasoning-delta"; delta: string }
+	| { type: "reasoning-end"; metadata?: Record<string, unknown> }
 	| {
 			type: "tool-call";
 			toolCall: {