ffsdeploy

OpenClawOpenClaw Configurator

Forge your agent configuration

Start from a Template

Pick a preset to auto-fill all sections, or start custom

LLM Provider & Model

Choose your primary AI model

Fallback Models

If the primary model fails, these are tried in order


Thinking Level

Controls how much the model reasons before responding


~1.5x avg — model decides when to think deeper

Agent Identity

Give your agent a name and personality


Display name shown in conversations

Character theme that shapes responses

Emoji avatar for the agent

Agent Behavior

Configure how your agent responds and who can access it


Users pair via a code before chatting

Only approved groups can interact

Uses the default lobster emoji prefix

Send complete messages at once

Show indicator immediately

Respond as fast as possible

Agent run stops after 600s

Auto-detect timezone

No authentication required

Messaging Channels

Select which platforms your agent connects to


Core Channels

Plugin Channels

Voice & Telephony

Enable phone call support with voice providers


Text-to-Speech Provider

Speech-to-Text Provider

Tool Profile & Built-in Tools

Control which tools your agent can use


Skills

Search ClawHub for pre-built skills to add to your agent


MCP Servers

Search the MCP Registry for tool servers to connect


Memory & Knowledge

How your agent remembers conversations and context


Session Policy

Control how conversation sessions are scoped and reset


Each user gets their own session

Reset after a period of inactivity

Session resets after 120 minutes of inactivity

Deployment

Choose how to install and where to host your agent

Install Method

Hosting Platform