Verify that:
- State is tracked when a run errors (getStateByRun returns error state)
- A new run after an error works correctly (runFinished flag resets,
second run tracked independently)
When the backend emits RunErrorEvent via the AG-UI protocol, several
components did not handle it:
- ProxiedCopilotRuntimeAgent: isRunning stayed true, causing
data-copilot-running to never transition to false (infinite spinner)
- StateManager: activeRun entries were never cleaned up, runFinished
flag never set (stale state on subsequent runs)
- Angular CopilotKitAgent: same isRunning bug as the proxy agent
- useAgentNodeName: node name stuck at last step instead of "end"
onRunErrorEvent is distinct from onRunFailed — the former handles
protocol-level RUN_ERROR events from the backend, the latter handles
local exceptions (network errors, deserialization failures).
## Summary
- **Root cause**: `CopilotChatMessageView` deduplicated messages with
`new Map(messages.map(m => [m.id, m]))`, which keeps only the *last*
occurrence of each ID. During streaming, when an agent fires multiple
tool calls in one turn, the same message ID appears multiple times:
first with text content, then with empty content + tool calls appended.
The last entry wins, wiping the text.
- **Fix**: Replace "keep last" with a **merge** strategy for assistant
messages — recover non-empty `content` from any earlier occurrence while
keeping the latest `toolCalls` (which accumulate). All other message
roles retain "keep last" behavior.
- **Test**: Added a regression test covering the three-occurrence
scenario (text → first tool call → second tool call) to confirm the
merged message renders both the original text and the final tool call
set.
## Test plan
- [x] New test `"preserves assistant text content when later duplicate
has empty content (multi-tool-call scenario)"` passes
- [x] Existing dedup tests (single duplicate, order preservation,
activity rendering) all pass unchanged
- [x] Full `@copilotkit/react-core` test suite: **950 passed, 0 failed**
(baseline was 949)
Closes#3470🤖 Generated with [Claude Code](https://claude.com/claude-code)
TanStackChatMessage.content uses any[] for multimodal so
messages are directly passable to any adapter without casts.
Add env.d.ts with vite/client types for CSS imports.
- Fix convertInputToTanStackAI silently dropping multimodal
content (images, audio, video, documents) by converting
AG-UI content parts to TanStack AI ContentPart format
- Switch react-router example from Hono server to a React
Router resource route using createCopilotRuntimeHandler
- Both agents now use BuiltInAgent factory pattern
- Remove hono/react-router-hono-server deps from example
The ui/message handler added messages to the chat but never called
runAgent(), so the agent never processed MCP-sent messages.
- Use copilotkit.runAgent({ agent }) through RunHandler for frontend
tools, context, tool execution, and abort support
- Send JSON-RPC response immediately after addMessage, before agent run
- Agent run is fire-and-forget with error logging
- followUp parameter: true=always, false=skip, default=user messages only
Fixes#3216
- Fix JSDoc: "while keeping the latest toolCalls" was wrong for the ?? case;
now says "recovers toolCalls from earlier occurrences if the latest is
undefined" and notes that [] is treated as intentional
- Add @internal annotation to signal export is for testing only
- Remove redundant AssistantMessage casts inside the role-narrowed branch
- Add missing content assertion to "uses latest content" render test
- Add test: [] toolCalls from later chunk is kept (not fallen back from)
- Add test: undefined content on both sides is handled without error
- Add changeset for @copilotkit/react-core patch
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
The base class @ag-ui/langgraph creates an Observable that calls
runAgentStream() but ignores the returned Promise. Any async rejection
becomes an unhandled promise rejection that crashes the process.
Override run() to properly route async errors through the Observable
error channel so downstream catchError handlers can process them.
The thread naming creates an additional run, using a thread name that
was something like
"thread-naming:${originalthreadiduuid}:${newrandomuuid}" which
shockingly is not a valid uuid. This just drops the first 2 parts and
keeps the random uuid so langgraph can stop being mad.
Keep both defaultThrottleMs (our branch) and inspectorDefaultAnchor +
design-skill context registration (main) in CopilotKitProvider.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
- Core setter rejects invalid values (NaN, Infinity, negative) instead of
storing them, preventing garbage from reaching downstream consumers.
- Provider initializes defaultThrottleMs synchronously during instance
creation so child hooks see the correct value on their first render.
- Remove phantom notificationThrottle JSDoc reference (API does not exist).
- Remove dead "default" branch in error source detection.
- Fix misleading @default comments re: 0-vs-undefined semantics and
cascade direction.
- Replace React-specific "re-renders" language in framework-agnostic core.
- Add test for dynamic provider defaultThrottleMs changes.
- Add tests for core setter validation behavior.
- Add defaultThrottleMs to renderWithCopilotKit test helper.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Replace @a2ui/lit with @a2ui/web_core 0.9, introduce createCatalog() API for
custom component catalogs, add basic (18 components) and minimal (5 components)
built-in catalogs, CSS variable theming for dark mode, eliminate XSS vector
(dangerouslySetInnerHTML removed from Text component).
Addresses Tyler's review comment on #3657 — defaultThrottleMs is configuration
that belongs on the copilotkit instance, not as a sibling field in the React
context value.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Jordan Ritter