feat(examples): ag-ui demo emits the protocol's SUBAGENT_* events

apps/website/content/blog/2026-08-27-langgraph-subgraphs-when-to-split.mdx

@@ -182,11 +182,14 @@ Its module docstring says so outright:
 
 ```text
 Mirrors cockpit/chat/subagents' orchestrator + `task` tool + `_run_subagent`
-structure, but each dispatch emits `subagent_activity` CUSTOM events
+structure, but each dispatch emits `subagent_activity` CUSTOM events [...]
+The backend's SubagentEmittingAgent expands those CUSTOM events into the
+protocol's standard SUBAGENT_STARTED / TEXT_MESSAGE_* (attributed via
+subagentRunId) / SUBAGENT_FINISHED / SUBAGENT_ERROR events
 ```
 
-The thing that differs is the transport: AG-UI's already carries a first-class delegation event.
-So the specialists stayed a flat `async` helper and progress reaches the frontend as a custom event dispatched from the tool body.
+The thing that differs is the transport: AG-UI already carries first-class delegation events — `SUBAGENT_STARTED`, `SUBAGENT_FINISHED`, and content events attributed to a child run.
+So the specialists stayed a flat `async` helper, the tool body dispatches its progress as custom events, and a thin wrapper on the server expands those into the protocol's standard subagent events on the wire.
 
 The subgraph was never required by the feature.
 It was required by the transport.
@@ -209,7 +212,7 @@ A node is already a unit.
 
 When the child really is a different graph — and the repo has exactly one of those, which is the case I owe you after arguing the other side this whole time.
 
-Our `examples/ag-ui` demo runs on that same AG-UI transport, and it emits the same `subagent_activity` events from the tool body.
+Our `examples/ag-ui` demo runs on that same AG-UI transport, and its research tool reaches the frontend the same way: the protocol's standard `SUBAGENT_*` events, with the child's messages and its own `lookup` tool call attributed to the child run.
 So it is not buying observability.
 It already had it.
 It compiles a child graph anyway.

apps/website/content/blog/2026-08-31-what-changes-when-the-runtime-changes.mdx

@@ -224,8 +224,8 @@ Our LangGraph subagent tracker is 543 lines.
 It infers subagent identity from stream namespaces, correlates namespaces back to tool-call ids, and requires you to configure `subagentToolNames: ['task']` in the provider so it knows which tool calls are delegations.
 That is client-side inference of a server-side fact, and inference is exactly as reliable as it sounds.
 
-AG-UI ships `ACTIVITY_SNAPSHOT` and `ACTIVITY_DELTA` as first-class events.
-The server declares "this is a subagent, here is its type, here is its status, here is its content."
+AG-UI ships `SUBAGENT_STARTED`, `SUBAGENT_FINISHED`, and `SUBAGENT_ERROR` as first-class events, and every content event can carry a `subagentRunId`.
+The server declares "this is a subagent, here is its name, here is the tool call that dispatched it, and these messages and tool calls belong to it."
 Our reducer projects those onto the neutral `Subagent` contract, and the AG-UI provider config for the subagents demo needs no subagent option at all.
 
 Declared beats inferred.

examples/ag-ui/angular/e2e/subagent-card.spec.ts

@@ -26,9 +26,9 @@ interface SubagentProbe {
 // each carrying `toolCallIds`/reasoning) and `toolCalls()` (the child's own
 // `lookup` calls, rendered as <chat-tool-call-card>). We read the projected map
 // directly rather than scraping the rendered card: it IS the data the card
-// renders, and asserting on it proves the ACTIVITY snapshot/delta pipeline
-// reconstructed the full reason→tool→answer transcript and settled to
-// `complete`, independent of card layout.
+// renders, and asserting on it proves the SUBAGENT_* + subagentRunId-attributed
+// event stream reconstructed the full reason→tool→answer transcript and
+// settled to `complete`, independent of card layout.
 async function readSubagents(page: Page): Promise<SubagentProbe> {
   return page.evaluate(() => {
     const ng = (window as unknown as { ng?: { getComponent?: (el: Element) => unknown } }).ng;
@@ -64,10 +64,12 @@ async function readSubagents(page: Page): Promise<SubagentProbe> {
 // `research` tool, the langgraph child subgraph runs a genuine reason → tool →
 // answer loop (an LLM call that returns a `lookup` tool_call, the offline
 // `lookup` tool, then a second plain LLM call that writes the summary), and the
-// AG-UI server converts the subagent_activity CUSTOM events into native
-// ACTIVITY_SNAPSHOT/ACTIVITY_DELTA. The @threadplane/ag-ui reducer projects the
-// activity to agent.subagents() (the ordered transcript chat-subagent-card
-// renders) and the child's research text must stay OUT of the parent's bubble.
+// AG-UI server's SubagentEmittingAgent expands the subagent_activity CUSTOM
+// events into the protocol's SUBAGENT_STARTED/FINISHED plus subagentRunId-
+// attributed TEXT_MESSAGE_* / TOOL_CALL_* events. The @threadplane/ag-ui
+// reducer projects them to agent.subagents() (the ordered transcript
+// chat-subagent-card renders) and the child's research text must stay OUT of
+// the parent's bubble.
 test('research delegation reconstructs the multi-message subagent transcript', async ({
   page,
 }) => {