Add AG-UI Streams to LangGraph Threads

Add Intelligence to your existing app#

Give users reconnection, catch-up, and delivery across devices around your LangGraph agent. Follow the Intelligence quickstart to connect your existing CopilotKit app and Runtime, then verify that a new conversation is saved. This enables AG-UI streams for runs through CopilotKit; no historical import is required.

Keep LangGraph or LangSmith storage and analytics configured for agent context and execution state. Maintain a stable mapping between CopilotKit threadId values and native LangGraph conversation identifiers so later runs reach the same conversation. See how streams work with framework storage.

Include earlier conversations (optional)#

To let users reopen conversations recorded before Intelligence was connected, use the importer below to copy supported historical content. Importing does not enable live delivery or establish ongoing database replication, and it cannot recover history the source no longer exposes.

The LangGraph importer reads LangGraph Server, LangGraph Platform, or LangSmith Deployment threads. This works for LangGraph Python, LangGraph TypeScript, and LangGraph FastAPI when the source history is available through the LangGraph SDK thread and run APIs. It does not import classic LangChain message history stores, LangSmith traces, or embedded checkpointers that are not exposed through LangGraph Server / Platform.

The following prerequisites and commands apply only to that historical import. Use Threads Drawer or Headless Threads to open imported and new conversations through the same UI.

Prerequisites#

• A CopilotKit app created with the CLI and CopilotKit Intelligence enabled.
• A LangGraph Server, LangGraph Platform, or LangSmith Deployment URL.
• A LangGraph API key, LangSmith API key, or LangChain API key accepted by the LangGraph SDK.
• Source threads that belong to one graph_id per thread. Threads spanning multiple graph IDs are skipped instead of imported partially.

Confirm the target project#

The importer targets the CopilotKit Intelligence project selected when you created the app with the CopilotKit CLI. If that is the project that should receive the LangGraph threads, continue to source configuration.

To target a different cloud-hosted project, select it before the dry run:

npx copilotkit@latest project select

The command updates the project selected for the current directory and writes its project-scoped runtime key to the app's generated .env.

Configure the source#

Set the LangGraph deployment URL and API key in your shell. The importer reads source credentials from framework-native environment variables, not raw CLI flags.

export LANGGRAPH_API_URL="https://your-langgraph-deployment"
export LANGGRAPH_API_KEY="..."

LANGSMITH_API_KEY or LANGCHAIN_API_KEY can be used instead when that is how your deployment authenticates. If your LangGraph Cloud workspace requires a tenant header, set LANGGRAPH_TENANT_ID or LANGSMITH_TENANT_ID.

Run a dry run#

Start with a preview. This reads the source, discovers graph IDs, counts conversations, reports skips, and estimates upload size without writing to CopilotKit Intelligence. It does not need a CopilotKit Intelligence URL or API key.

npx copilotkit@latest import --source langgraph --dry-run

Map graph IDs to CopilotKit agent IDs#

The source agent key for LangGraph is the thread's graph_id. Map each discovered graph ID to the agentId your live CopilotKit runtime uses.

{
  "support_graph": "support-agent",
  "research_graph": "research-agent"
}

Using the same agentId as live traffic keeps imported history and future conversations grouped together.

Prepare the CopilotKit Intelligence destination#

A real import needs the destination app-api URL and project-scoped runtime key. A CLI-created starter writes them to .env, but the importer reads the current process environment and does not load .env or .copilotkit/project.json automatically.

Copy the generated values into your shell before importing:

export INTELLIGENCE_API_URL="https://..."
export CPK_INTELLIGENCE_API_KEY="cpk-..."

COPILOTKIT_API_KEY is also accepted for the key. You can pass the destination directly with --api-url and --api-key instead.

Import the threads#

Run the import after the dry run and agent mapping look right.

npx copilotkit@latest import \
  --source langgraph \
  --agent-map ./agent-map.json \
  --user-source configurable.user_id

Use --user-source to tell the importer where to find the end-user id for each thread. You can also pass a literal when every imported thread belongs to the same user:

npx copilotkit@latest import \
  --source langgraph \
  --agent-map ./agent-map.json \
  --user-source literal:imported-user

For self-hosted CopilotKit Intelligence, pass the target connection explicitly:

npx copilotkit@latest import \
  --source langgraph \
  --api-url "$INTELLIGENCE_API_URL" \
  --api-key "$CPK_INTELLIGENCE_API_KEY" \
  --agent-map ./agent-map.json \
  --user-source configurable.user_id \
  --yes

Verify the imported threads#

Open the Threads Drawer, select an imported LangGraph conversation, and confirm that its thread history appears in the chat.

Re-run or replace#

Re-running the same import is idempotent. Conversations already imported from the same source thread are skipped.

Use --replace to delete and re-import previously imported threads:

npx copilotkit@latest import \
  --source langgraph \
  --agent-map ./agent-map.json \
  --user-source configurable.user_id \
  --replace

Continue using LangGraph persistence#

Importing copies supported history; it does not establish ongoing database replication. Intelligence rename, archive, and delete operations affect only Intelligence records, leaving the native LangGraph thread records unchanged.

Your CLI-created app sends future CopilotKit conversations to CopilotKit Intelligence. To keep each conversation continuous, reopen it with the same CopilotKit threadId. If your agent remains wired to a durable LangGraph checkpointer or Platform deployment, the same future runs continue to persist in LangGraph or LangSmith for native storage and analytics.

• Threads Drawer: already included in CLI-created starters. Use the Threads Drawer guide to customize its ready-made thread UI.
• Headless Threads: use the Headless Threads guide only when you need a custom UI. Select a thread with useThreads, store its thread.id, and pass that value to your chat component as threadId.

If your backend maps CopilotKit thread IDs to LangGraph thread IDs, keep that mapping stable so a resumed CopilotKit thread reaches the expected LangGraph thread. LangGraph Platform thread IDs must be UUIDs, so persist a mapping instead of reusing the CopilotKit threadId when necessary.

Create one new conversation through CopilotKit, then confirm that it appears in CopilotKit Intelligence and in the LangGraph or LangSmith persistence store that remains connected to your agent.