Qoren documentation for managed agent hosting.

1. Choose a runtime

Start by choosing OpenClaw or Hermes. Use OpenClaw when you need OpenClaw compatibility or an existing OpenClaw setup. Use Hermes when you want Qoren's native configuration and operations workflow.

2. Configure the agent

Define what the agent is supposed to do before giving it credentials.

• Persona and operating instructions.
• Tasks, schedules, and event triggers.
• Allowed tools and integration boundaries.
• Workspace files, persistent state, and expected outputs.

3. Connect model access

Qoren can use managed model access or bring-your-own-key on supported plans. Choose the model policy before a production deployment so spend controls and provider ownership are clear.

4. Add secrets carefully

Add only the credentials an agent needs. Prefer least-privilege tokens, separate agent accounts, and credentials that can be rotated without changing unrelated systems.

5. Deploy and monitor

Deploy the agent to a managed environment, then monitor activity, logs, usage, and spend. Review the first runs before allowing broader autonomy.

6. Decide when managed hosting is not right

Use local self-hosting when you need offline execution, a free local experiment, or full host-level control. Use Qoren when the agent should stay online and you do not want to maintain the infrastructure.