A minimal-code conversational agent builder built on the Agent Development Kit, representing the evolution of Dialogflow CX.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install cx-agent-studio
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install cx-agent-studio using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is CX Agent Studio?
CX Agent Studio is the successor to Dialogflow CX, offering a streamlined, minimal-code environment for building sophisticated conversational agents. By utilizing Openclaw Skills, developers can harness the power of Gemini to generate agents from a few sentences, creating responsive and intelligent interfaces. The platform focuses on low-latency interactions and bi-directional streaming, making it ideal for both text-based and voice-based customer experience automation. It operates on the Agent Development Kit (ADK) to bridge the gap between high-level intent and technical execution.
CX Agent Studio Use Cases
- Automating customer service workflows with AI-driven root and sub-agent architectures
- Developing ultra-low latency voice assistants for real-time customer engagement
- Migrating legacy Dialogflow CX flows into modern, LLM-powered agent environments
- Implementing secure tool-calling for dynamic backend integrations during conversations
- Creating complex evaluation suites for regression testing and persona simulation
How CX Agent Studio Works
- Generate an initial agent structure by providing a clear 1-2 sentence goal to the AI-augmented builder.
- Configure knowledge documents and tools to ground the agent responses in specific organizational data.
- Structure agent instructions using specific syntax and XML tags to guide behavior and persona.
- Define delegation logic where a root steering agent orchestrates tasks across specialized sub-agents.
- Implement Python-based callbacks to handle advanced logic, state validation, or custom JSON payloads.
- Run comprehensive evaluations using Scenario and Golden test cases to ensure reliability and performance.
CX Agent Studio Setup
To get started with this skill within the Openclaw Skills framework, ensure your environment is configured for the Agent Development Kit (ADK). Begin by initializing your agent project and preparing any existing knowledge documents:
# Initialize your local environment for CX Agent Studio development
mkdir cx-agent-project && cd cx-agent-project
# Organize your documentation for knowledge grounding (must be < 8MB total)
cp /path/to/faqs.pdf ./knowledge/
Ensure you have defined your root agent goal before using the generation tools.
CX Agent Studio Data Schema & Taxonomy
The skill organizes data around several key entities to ensure a structured conversational flow and reliable metadata taxonomy:
| Component | Description | Syntax Reference |
|---|---|---|
| Agents | Root or sub-agents orchestrating specific tasks | {@AGENT: Agent Name} |
| Tools | Connections to external APIs and Python scripts | {@TOOL: tool_name} |
| Variables | Runtime conversation data and session storage | {variable_name} |
| Instructions | XML or Markdown formatted guidance for model logic | references/instructions.md |
| Evaluation | Golden and Scenario test cases for quality assurance | references/evaluation.md |
| Callbacks | Python hooks for fine-grained execution control | references/callbacks.md |
CX Agent Studio Advanced Features
- Bi-directional streaming for low-latency, natural-sounding voice interactions
- Asynchronous tool calling to maintain conversation continuity while fetching backend data
- Advanced Python hooks for before_agent_callback and after_model_callback execution control
- Automated Scenario Test Cases that simulate diverse user goals and personas for stress testing
- Enterprise-grade guardrails for prompt attack protection and responsible AI policy enforcement
- Deep integration with legacy Dialogflow CX flows for deterministic business logic validation
SKILL.md
Loading
Related Openclaw Skills
WaveStreamer AI Forecasting
An AI-agent-exclusive forecasting platform for predicting industry milestones and earning points through evidence-based reasoning.
Polymarket Latest Events
A high-performance tool for fetching live prediction market events, volume, and real-time betting odds from Polymarket.
VeADK Agent Generator
A powerful tool for generating VeADK Agent code from scratch or converting existing Langchain and Dify workflows into the VeADK framework.
Content Strategy Analyzer
An automated toolkit for crawling websites to generate localized content strategies, keyword research, and competitor analysis reports.
Dialogflow CX Advanced
A powerful integration for managing advanced Google Dialogflow CX infrastructure, including environments, webhooks, and automated deployment stages.
Dialogflow CX Agent Management
A robust automation interface for managing Google Dialogflow CX chatbot agents via the v3beta1 REST API.
Featured*
