Intent Engine is an AI-driven intent recognition and intelligent routing system with a built-in Web management dashboard, providing accurate prompt classification for multi-agent workflows.
The fastest way to install a skill directly from the registry.
npx clawhub@latest install intent-engine
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
Copy this prompt to OpenClaw to install it automatically.
Help me install intent-engine using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Get the raw skill files in a ZIP archive.
Intent Engine is a robust, lightweight Python-based routing system designed to analyze, classify, and redirect user prompts to the correct AI execution path. By utilizing a hybrid scoring algorithm that combines keyword matching, regular expressions, priority weighting, and text-length penalties, it dynamically categorizes inputs into four primary domains: Code (CODE), Knowledge (KNOW), Task (TASK), and Chat (CHAT).
Integrating this tool with Openclaw Skills allows developers to build highly responsive, multi-agent architectures that process queries intelligently. The inclusion of a comprehensive Web UI dashboard provides real-time testing, performance metrics, and hot-reloadable intent management without requiring system restarts.
To set up and run Intent Engine within your Openclaw Skills ecosystem, follow these simple installation steps:
# Navigate to the intent-engine directory
cd intent-engine
# Install the required dependencies
pip install -r requirements.txt
# Run the server
python run.py
Once running, open your browser and navigate to http://localhost:5700 to access the visual management dashboard.
Intent Engine manages its data and configuration seamlessly through localized storage.
| Path | Description |
|---|---|
intent_engine/engine.py |
Core classification logic and scoring algorithm |
intent_engine/storage.py |
JSON-based persistent storage handler |
intent_engine/models.py |
Structured Python data models for intents |
intent_engine/api.py |
Flask REST API endpoints |
static/index.html |
Frontend UI assets |
data/intents.json |
Hot-reloadable persistent configuration (auto-generated) |
#3b82f6): Software engineering, testing, review, refactoring, and debugging tasks.#8b5cf6): Information inquiries, tutorials, comparisons, conceptual queries, and suggestions.#f59e0b): Automation scripts, file systems operations, shell command executions, and task scheduling.#10b981): Greeting, conversational dialogue, clarification prompts, and social pleasantries.* 4.0) with adjustable weighted keyword scoring (* 2.0 * weight) and static priority offsets (* 0.05).skill-creator, workflow-verifier, or mempalace-assistant.Loading
An AI-powered wellness protocol agent designed to guide users through Anthony William's 'Medical Medium Cleanse to Heal' framework, providing precise detox and dietary strategies.

An advanced financial analysis engine that generates interactive, 14-module HTML evaluation reports on corporate value, competitive moats, and absolute DCF valuation.

An interactive AI agent skill that delivers life-saving philosophy, recovery lessons, and resilience frameworks from Jeremy Renner's memoir.

An AI agent skill encapsulating Archbishop Desmond Tutu's definitive philosophy on South Africa's Truth and Reconciliation Commission, Ubuntu, and restorative justice.

An executable knowledge toolkit based on Adam Higginbotham's "Midnight in Chernobyl" that enables AI agents to analyze the design flaws, human cost, and cover-up of the 1986 disaster.

An AI agent skill that translates Mark Manson's bestselling framework into an interactive, vulnerability-based dating and self-improvement assistant.








































