Deliver OpenClaw observability data to Agent-lens and All-in-One Observability - ApsaraDB for ClickHouse

openclaw-exporter-to-langfuse is an observability extension plugin for OpenClaw. It delivers Agent runtime data to the Agent-lens and All-in-One Observability platforms of ApsaraDB for ClickHouse, helping you comprehensively monitor and analyze the running status of your Agent applications.

Observability capabilities

openclaw-exporter-to-langfuse provides two complementary observability capabilities:

Capability	Data pipeline	Delivery component	Primary use case	Viewing tool
Deliver to Agent-lens	OpenClaw → openclaw-exporter-to-langfuse → Langfuse	openclaw-exporter-to-langfuse	Analyze individual Agent call chains and LLM interaction results	Langfuse
Deliver to All-in-One Observability	OpenClaw → diagnostics-otel → otelcol-contrib → ClickHouse	otelcol-contrib	Analyze OpenClaw runtime metrics, Gateway logs, and overall health status	HyperDX

Open source repository: https://github.com/aliyun/openclaw-exporter-to-langfuse

Limits

The OpenClaw plugin openclaw-exporter-to-langfuse requires Node.js 18 or later.
To deliver traces, metrics, and logs to All-in-One Observability, OpenClaw must support the diagnostics-otel plugin. We recommend that you use OpenClaw v2026.2.19 or later.
otelcol-contrib listens on ports 4317 and 4318 by default. Make sure these ports are not occupied.

Prerequisites

An ApsaraDB for ClickHouse cluster is created, and the following information is obtained:
- VPC endpoint
- Username (for example, openclaw, which must have read and write permissions)
- Password
All-in-one observe suite is enabled for the ClickHouse cluster in the console.
The Agent Observability Suite (Agent-lens) is enabled for the ClickHouse cluster in the console. A project is created and the following information is obtained:
- VPC endpoint
- Public Key (in the format of pk-lf-xxx)
- Secret Key (in the format of sk-lf-yyy)
OpenClaw is installed and running.

Considerations

Traces, metrics, and logs are different observability dimensions. By default, traces are delivered to Agent-lens, while metrics and logs are delivered to All-in-One Observability.
To analyze traces in All-in-One Observability, specify the --diag-traces true parameter during installation.
The time zone of the ClickHouse cluster must be set to UTC. Otherwise, the Agent Observability Suite may display data anomalies.

Procedure

One-click installation through OpenClaw

We recommend that you use the OpenClaw conversational interface to complete the installation. OpenClaw automatically identifies the local configuration path, installation directory, and runtime environment.

Have OpenClaw download the installation instructions

Enter the following content in the OpenClaw dialog box:

Download the installation file (overwrite if it already exists): https://ck-langfuse-public.oss-cn-beijing.aliyuncs.com/openclaw-exporter-to-langfuse/INSTALLATION.md, and follow the prompts to install.

OpenClaw returns the installation parameter instructions.

Provide installation parameters and run the installation

Enter the following content in the OpenClaw dialog box, and replace the parameters with actual values:

Install both openclaw-exporter-to-langfuse and otelcol-contrib.

openclaw-exporter-to-langfuse configuration:
endpoint = http://<LANGFUSE_ENDPOINT>:3000/api/public/otel/v1/traces
pk = <PK>
sk = <SK>
serviceName = <SERVICE_NAME>
debug = true

otelcol-contrib configuration:
ck-endpoint = http://<CK_ENDPOINT>:8123
username = <USERNAME>
password = <PASSWORD>

The following table describes the parameters.

Parameter	Description
`endpoint`	The Langfuse trace endpoint. Replace `<LANGFUSE_ENDPOINT>` with the VPC endpoint of Agent-lens.
`pk`	The public key of the Agent-lens project.
`sk`	The secret key of the Agent-lens project.
`serviceName`	The name of the Agent service, used to identify the data source.
`debug`	Specifies whether to enable debug logging for the plugin. We recommend that you set this parameter to `true` during initial installation or troubleshooting.
`ck-endpoint`	The endpoint of the ClickHouse cluster. Replace `<CK_ENDPOINT>` with the VPC endpoint of the cluster.
`username`	The username for the ClickHouse cluster.
`password`	The password for the ClickHouse cluster.

After OpenClaw returns a success message, both openclaw-exporter-to-langfuse and otelcol-contrib are configured.

View the component configuration
Enter the following content in the OpenClaw dialog box to confirm the installation result:
```
View the configuration of both components
```
OpenClaw returns the installation status and configuration information of both components. The expected output includes:
- openclaw-exporter-to-langfuse (Langfuse plugin): Shows enabled as true, along with the configured values of endpoint, authorization, serviceName, and other parameters.
- diagnostics-otel + otelcol-contrib (All-in-One Observability): Shows the enabled status and endpoint of diagnostics-otel, along with the running status, gRPC/HTTP receiver ports, and ClickHouse endpoint information of the otelcol-contrib service.
If both components are shown as enabled with the correct parameters, the installation is successful.

Verify data synchronization

View traces through the Agent Observability Suite

Log on to the Agent Observability Suite console.
Go to the target project.
Open the Traces page.
Initiate an Agent conversation in OpenClaw.
Check whether a new trace record is generated.

If the configuration is correct, new trace records appear in the Traces list. Click a trace to view the complete call chain tree (including nodes such as enter_openclaw_system, invoke_agent_main, and tool_call), along with the latency, token consumption, input/output content, and LLM interaction details for each node.

Verify data through ClickHouse

Log on to ClickHouse and run the following SQL statements to verify that data is successfully written:

SHOW DATABASES;

SHOW TABLES FROM clickobserve_service;

SELECT count() FROM clickobserve_service.otel_logs;

SELECT count() FROM clickobserve_service.otel_metrics_sum;

View and analyze data through All-in-One Observability

Log on to the ClickHouse All-in-One Observability console.
Analyze data by using AI Notebook.

The following is an example of a common analysis query:

Show OpenClaw error logs from the past 1 hour

AI Notebook automatically analyzes the query intent, generates and executes the corresponding SQL statements, and returns analysis results including the time, source, and content of error logs.

FAQ

Traces are not visible in All-in-One Observability

By default, traces are delivered only to Langfuse (Agent-lens) and not to ClickHouse All-in-One Observability. To analyze traces in All-in-One Observability, reinstall the plugin with the following parameter:

--diag-traces true