SoCLaaS

Home / Setup Guides / OpenClaw

Agent Harness

Connect OpenClaw to SoCLaaS

Give an OpenClaw agent harness an OpenAI-compatible model endpoint backed by SoCLaaS, so multi-step tasks run against your gateway key and model.

No generated config for OpenClaw. The portal's Client Setup page provides ready-made configs for Codex, OpenCode, and Qwen Code. For OpenClaw, use the manual connection values below.

OpenClaw's configuration file name and field names vary by release. Use the connection values below in your version's OpenAI-compatible provider block, then verify against the OpenClaw documentation.

Connection values

SettingValue
Provider typeOpenAI-compatible
Base URLhttps://soclaas-api.comp.nus.edu.sg/v1
API keyFrom the SoCLaaS portal, stored in SOCLAAS_API_KEY
ModelA chat-capable ID from GET /v1/models

Steps

  1. Get an API key

    Sign in to the SoCLaaS portal from the NUS network and create a personal key.

  2. Export the key

    export SOCLAAS_API_KEY="YOUR_API_KEY"
  3. Find a model ID

    curl https://soclaas-api.comp.nus.edu.sg/v1/models \
      -H "Authorization: Bearer $SOCLAAS_API_KEY"

    OpenClaw harnesses usually need a model that supports tool calling. Pick a chat-capable ID and note it for the next step.

  4. Open the provider settings

    Locate the OpenAI-compatible or custom provider section in your OpenClaw configuration. Fill in the base URL, API key reference, and model ID from the table above.

    If OpenClaw reads keys from the environment, reference SOCLAAS_API_KEY. If it stores keys inline, paste the key directly and keep the file out of version control.

  5. Test the connection

    Run a single-step task first, then a task that requires a tool call. This confirms both chat completions and tool support.

Troubleshooting