Ship the batch: undo, parallel subagents, lazy MCP, hot-reloaded skills; docs and CI/CD

Loop and ergonomics batch across Now/Next and Maintenance:
- /undo and /redo via pre-prompt file snapshots (snapshot.ts)
- task takes a tasks[] array and runs investigations concurrently (subagent.ts)
- lazy MCP tools: mcp_list/mcp_inspect/mcp_call meta-tools, eager opt-in (mcp.ts, config.ts)
- skill tool reads its list live so a mid-session install is callable next turn (skills.ts)
- tool-name lists (tool-kinds.ts) derived from a mutating() marker; gates previously ungated writes
- prune/session recovery path summarized, and step-back doom-loop primitive (step-back.ts)
- @file completion re-walks on a slow cooldown; estimateTokens and pricing labeled as estimates

Docs: README, CHANGELOG, docs/{mcp,architecture,development} updated to match.
CI/CD: bun install-store cache and concurrency gates on both workflows; release.yml now
composes file-based release notes via scripts/make-release-notes.ts and verifies every binary.

test/helpers.ts

@@ -1,4 +1,10 @@
-import type { LanguageModelV4Usage } from '@ai-sdk/provider';
+import { simulateReadableStream } from 'ai/test';
+import type {
+  LanguageModelV4GenerateResult,
+  LanguageModelV4StreamPart,
+  LanguageModelV4StreamResult,
+  LanguageModelV4Usage,
+} from '@ai-sdk/provider';
 import type { AppHooks } from '../src/ui/App';
 
 /**
@@ -15,6 +21,46 @@ export function usageOf(input: number, output = 1): LanguageModelV4Usage {
   };
 }
 
+/**
+ * A `doStream` return value built from `simulateReadableStream` chunks.
+ *
+ * Written out as `{ stream: simulateReadableStream(...) } as any` in most UI test
+ * files, because the SDK's `ReadableStream` and `simulateReadableStream`'s type do
+ * not unify. This is the one place that difference is absorbed.
+ */
+export function streamOf(chunks: LanguageModelV4StreamPart[]): LanguageModelV4StreamResult {
+  return {
+    stream: simulateReadableStream({
+      chunks,
+      chunkDelayInMs: null,
+      initialDelayInMs: null,
+    }),
+  } as unknown as LanguageModelV4StreamResult;
+}
+
+/** The text and finish chunks that make a stream say one thing and stop. */
+export function textChunks(body: string, usage: LanguageModelV4Usage = usageOf(1)): LanguageModelV4StreamPart[] {
+  return [
+    { type: 'text-start', id: '0' },
+    { type: 'text-delta', id: '0', delta: body },
+    { type: 'text-end', id: '0' },
+    { type: 'finish', finishReason: { unified: 'stop', raw: 'stop' }, usage },
+  ];
+}
+
+/**
+ * A `doGenerate` return value. `warnings` is required by the SDK type but empty in
+ * every test, so it is filled here rather than repeated at each call site.
+ */
+export function generateResult(body: string, usage: LanguageModelV4Usage = usageOf(1)): LanguageModelV4GenerateResult {
+  return {
+    content: [{ type: 'text', text: body }],
+    finishReason: { unified: 'stop', raw: 'stop' },
+    usage,
+    warnings: [],
+  };
+}
+
 /** Default AppHooks for UI tests; override only what a test cares about. */
 export function testHooks(over: Partial<AppHooks> = {}): AppHooks {
   return {