OpenClaw FAQs

How Do I View the Number of Tokens Consumed by Huawei Cloud MaaS AI Models?

If you are using Huawei Cloud MaaS AI models, perform the following operations to view the number of consumed tokens. If you are using AI models from other providers, check with the respective model provider.

Method 1: On the FlexusL console

1. Log in to the FlexusL console and click Configure Now on a resource card to go to the application details page.
2. Click Call Details on the right of the target model to view call details on the Huawei Cloud MaaS console, including the total number of calls, failed calls, tokens, input tokens, output tokens, and end-to-end latency. For more information, see Viewing the Call Data and Monitoring Metrics of Real-Time Inference in MaaS.

Method 2: On the MaaS console

Log in to the MaaS console. In the navigation pane on the left, choose Management & Statistics > Calls to view the call statistics of the enabled models in the service list.

Huawei Cloud MaaS models configured for the OpenClaw image on FlexusL instances are in the CN-Hong Kong region. Ensure that you have switched to the CN-Hong Kong region.

Huawei Cloud MaaS models are billed based on the actual usage after being enabled. For more billing details, see Billing Items for MaaS Built-in Model Services.

What Should I Do If the Gateway Status Is Unknown or Unhealthy?

The initial status of the OpenClaw gateway is Loading. If the status changes to Unknown or Unhealthy after several minutes, click Restart. The restart takes some time. After the gateway is restarted successfully, the status changes to Running.

If the gateway status is still Unknown or Unhealthy after the restart, use the following methods to check the cause.

• Method 1: Check whether the gateway status is normal.
  Remotely log in to the FlexusL instance and query the gateway status. If an error is reported, fix it based on the error message.

  su - openclaw
  openclaw gateway status

  For example, the following error message indicates that the configuration is invalid.

  

  Modify the configuration and query the gateway status again. After the gateway status becomes normal, return to the FlexusL console and refresh the gateway status.

  vim /home/openclaw/.openclaw/openclaw.json

  

  

  • A FlexusL instance does not have an initial password. After creating a FlexusL instance, you need to set a password before logging in to it. For details, see Resetting the Password for a FlexusL Instance.
  • For details about how to log in to a FlexusL instance, see Login Modes.
• Method 2: Check whether Cloud Operations Center (COC) has been enabled and authorized.

  Querying the OpenClaw gateway status depends on COC. You need to enable and authorize COC first. Log in to the COC console. If the content in the following figure is displayed, enable and authorize COC as prompted.

  

  If you have already enabled the COC agency but have not granted required permissions, choose Resources > Application and Resource Management in the left navigation pane and click Obtain Permission at the top of the page. After the permissions are obtained, click Synchronize Resource on the Flexus L Instance tab. Return to the FlexusL console and refresh the gateway status.

  

If the gateway status is still Unknown or Unhealthy after you perform the operations described above, submit a service ticket.

What Are the Common OpenClaw Commands?

Remotely log in to the FlexusL instance and switch to the OpenClaw user. Then, run other commands as needed.

su - openclaw

openclaw status

openclaw --help

openclaw gateway start

openclaw gateway stop

openclaw gateway restart

openclaw gateway status

openclaw skills list

openclaw skills list | grep ready

openclaw plugins list

openclaw plugins -h

openclaw logs

openclaw logs --follow

• A FlexusL instance does not have an initial password. After creating a FlexusL instance, you need to set a password before logging in to it. For details, see Resetting the Password for a FlexusL Instance.
• For details about how to log in to a FlexusL instance, see Login Modes.

How Do I Query the Skills Supported by OpenClaw?

• Method 1: On the FlexusL console, click the resource card to go to the application details page. In the Configure Skills (Optional) module, you can view all installed skills.
• Method 2: In the chat box with the bot, ask the bot what skills are available. The bot will automatically search for and sort the skills for you.
• Method 3: Remotely log in to the FlexusL instance and run the following commands:

  su - openclaw
  openclaw skills list

  

  • A FlexusL instance does not have an initial password. After creating a FlexusL instance, you need to set a password before logging in to it. For details, see Resetting the Password for a FlexusL Instance.
  • For details about how to log in to a FlexusL instance, see Login Modes.

What Should I Do If Skills Failed to Be Installed?

Symptom: When you install skills online on the FlexusL console by referring to Skill Installation, a message is displayed indicating that the ClawHub community is busy and the installation fails.

Possible cause: The installation failed due to rate limits on skill installations within the ClawHub community.

Solution: Use the ClawHub token to log in to ClawHub and then install the skills.

1. Visit the ClawHub community and log in using your GitHub account. If you do not have an account, create one as instructed.

   

2. Obtain the ClawHub token. Choose Settings from the username drop-down list and click Create token. Copy and save the token.

   

3. Remotely log in to the FlexusL instance and replace Token with the token obtained in the previous step.

   clawhub login --token Token

   

4. Run the command below. If your ClawHub username is displayed, the login is successful.

   clawhub whoami

   

5. Return to the FlexusL console and install skills again.

What Should I Do If a Skill Cannot Be Used Normally After Being Installed Online?

Symptom: After OpenClaw is deployed and a skill is installed online on the application details page, the skill cannot be used during conversations with the bot.

Possible cause: Some skills (such as Playwright MCP) require additional components to function. However, these dependencies are not automatically installed during skill installation. You need to install them manually.

Solution: Go to the ClawHub community Skills to check the skill description. If the skill has dependencies, install them before using the skill. For example, the Playwright MCP skill specifies in its skill.md that Chromium must be installed before first use. You can either send the command in the bot chat box to let the bot run the installation, or remotely log in to your FlexusL instance to execute the command manually.

Why Does Not the Bot Respond During a Conversation?

Symptom: The configuration has been completed as instructed, but the bot does not respond during a conversation.

Possible cause: The API key of the AI model is incorrect.

Solution: Change the API key of the AI model by referring to Model Configuration.

Why Am I Seeing "HTTP 401: invalid authorization header"?

Symptom: The configuration has been completed as instructed. When you chat with the bot, the bot replies "HTTP 401: invalid authorization header".

Possible cause: The API key of the AI model is incorrect.

Solution: Configure or change the API key of the AI model by referring to Model Configuration.

Why Am I Seeing "HTTP 403: input text may contain sensitive information, please try again"?

Symptom: The configuration has been completed as instructed. When you chat with the bot, the bot replies "HTTP 403: input text may contain sensitive information, please try again".

Possible cause: The input or output content in the historical chat may contain sensitive information.

Solution: Enter /new in the chat box to start a new conversation. Ensure that the new input does not contain sensitive information.

Why Am I Seeing "HTTP 404: not found"?

Symptom: The configuration has been completed as instructed. When you chat with the bot, the bot replies "HTTP 404: not found".

Possible cause: The base_url of the custom model is incorrect.

Solution: Custom models only support OpenAI- and Anthropic-compatible API addresses, for example, https://api.openai.com/v1. Obtain the API address from the model provider and change the base_url of the custom model by referring to Model Configuration.