OpenClaw GitHub: 162 Agent Templates You Can Deploy Today

The Largest Open-Source AI Agent Template Library

The awesome-openclaw-agents repository on GitHub contains 162 agent templates organized into 24 categories. It is, by a significant margin, the largest collection of ready-to-deploy AI agent configurations available anywhere as open source.

Why does this matter? Because starting an AI agent from scratch is slow. You need to define the agent's role, write its system instructions, specify its skills, set behavioral rules, and test iterations until it works reliably. A good SOUL.md file can take an hour or more to write from zero. With 162 templates, you skip most of that work. Find an agent that matches your use case, copy the SOUL.md, make a few adjustments, and you are running.

The repository has grown rapidly since its launch. It has accumulated over 360 GitHub stars, receives hundreds of unique visitors daily, and is actively maintained with new templates added regularly through both maintainer contributions and community submissions. The top referral sources include GitHub itself, Google search, Reddit, ChatGPT, and Twitter, which reflects the broad interest in practical, deployable agent configurations.

How Templates Work: The SOUL.md Format

Every template in the repository is built around one file: SOUL.md. This is the OpenClaw agent configuration format. It is plain markdown, human-readable, and defines everything about an agent in a single document. No JSON schemas, no YAML indentation headaches, no proprietary DSL. Just markdown.

A SOUL.md file contains several key sections that together define a complete agent:

The beauty of this format is its simplicity. You can read any SOUL.md file and immediately understand what the agent does. You can edit it in any text editor. And you can version-control it with Git just like any other code file. Full Agent OS templates also include AGENTS.md (for multi-agent orchestration), HEARTBEAT.md (for scheduled tasks), and WORKING.md (for runtime memory).

# Project Manager Agent

## Identity
You are a senior project manager who coordinates software
development teams. You break down projects into tasks,
assign priorities, track progress, and ensure deadlines
are met.

## Rules
- Always respond in English
- Never fabricate project status data
- Ask clarifying questions if requirements are vague
- Keep status updates concise and actionable

## Skills
- file: Read and write project documentation
- browser: Research technical topics and dependencies

## Workflow
1. Understand the project scope and objectives
2. Break down work into discrete tasks
3. Assign priorities (P0, P1, P2) based on impact
4. Track progress and flag blockers
5. Generate daily standup summaries

All 24 Categories: Complete Breakdown

The 162 templates are organized into 24 categories covering virtually every business function. Each category contains between 4 and 12 templates with varying levels of specialization. Here is the complete list with example agents from each category.

The diversity of categories reflects real-world demand. Software engineering and content writing have the most templates because those are the most common use cases for AI agents today. But specialized categories like Supply Chain, Compliance, and Voice are growing as teams discover that AI agents can handle domain-specific workflows that were previously too niche for generic automation tools.

Most Popular Templates by GitHub Traffic

GitHub traffic data reveals which templates get the most views. These are the agents people search for, click on, and deploy most often. Here are the top 10 templates by page views.

The Project Manager template (Orion) leads by a wide margin. This makes sense: project management is a universal need, and an AI agent that can break down tasks, track progress, and generate status reports saves hours every week. The Software Engineer and Content Writer templates round out the top three, reflecting the two largest user segments in the OpenClaw community.

Newest Categories: What Makes Them Unique

Five categories were added more recently based on community demand. Each one addresses a specialized domain where AI agents bring distinct value.

Supply Chain

Supply chain agents handle logistics coordination, demand forecasting, and supplier evaluation. The Logistics Coordinator template, for example, monitors shipment statuses, flags delays, and suggests rerouting options. The Demand Forecaster analyzes historical sales data and seasonal patterns to predict inventory needs. These templates are built for teams that need to process large volumes of operational data and make time-sensitive decisions.

Compliance

Compliance is tedious, repetitive, and high-stakes. The GDPR Auditor template scans data processing workflows and flags potential violations. The SOC2 Checklist Agent walks through the SOC2 control requirements and tracks which ones are met, partially met, or missing. The Regulatory Monitor watches for regulatory updates in specified jurisdictions and summarizes changes relevant to the business.

Voice and Conversational

Voice agents are designed for spoken interaction flows. The Voice Assistant template handles natural language voice commands and routes them to appropriate actions. The IVR Designer creates interactive voice response menus and call routing logic. The Conversation Flow Builder maps out dialog trees for customer-facing chatbots with branching logic and fallback handling.

Customer Success

Customer success agents focus on retention and expansion. The Churn Risk Analyzer monitors usage patterns and flags accounts that show signs of disengagement. The Onboarding Guide walks new customers through product setup with personalized recommendations. The NPS Survey Agent manages survey distribution, response collection, and sentiment analysis.

Automation

The Automation category contains general-purpose workflow agents. The Workflow Automator connects multi-step processes across different systems. The API Integration Agent handles authentication, data mapping, and error handling for third-party API connections. The Notification Dispatcher manages alert routing across Slack, email, Telegram, and other channels based on configurable rules.

How to Use a Template: 3 Methods

There are three ways to go from browsing a template on GitHub to running it as a live agent. Choose the method that fits your technical comfort level and deployment needs.

Method 1: Manual Copy

The most straightforward approach. Browse the repository, find a template you like, copy the SOUL.md content, and paste it into your own agent directory. This gives you full control over every line of the configuration.

# Create a new agent directory
mkdir my-project-manager && cd my-project-manager

# Copy the SOUL.md content from GitHub
# (or clone the repo and copy the file locally)

# Register the agent with OpenClaw
openclaw agents add project-manager ./

# Start the gateway and test
openclaw gateway start
openclaw agent --agent project-manager --message "Create a sprint plan"

Method 2: Deploy with CrewClaw

If you want a visual interface and a ready-to-deploy package, use CrewClaw. Browse all 162 templates in the agent gallery, pick one, customize it through the guided wizard, and download a complete deployment package. The package includes the configured SOUL.md, directory structure, skills setup, and deployment instructions. No terminal required for the selection and configuration steps.

Method 3: Fork the Repository

For teams that want to maintain their own curated collection of templates, fork the entire repository. This gives you all 162 templates locally, plus the ability to add your own, modify existing ones, and keep everything under version control. You can pull updates from the upstream repository to get new templates as they are added.

# Fork on GitHub, then clone your fork
git clone https://github.com/YOUR_USERNAME/awesome-openclaw-agents.git
cd awesome-openclaw-agents

# Browse categories
ls agents/

# Copy a template to your workspace
cp -r agents/project-management/project-manager/ ~/my-agents/pm/

# Stay in sync with new templates
git remote add upstream https://github.com/mergisi/awesome-openclaw-agents.git
git pull upstream main

Contributing Your Own Agent

The repository accepts community submissions. If you have built an agent that solves a real problem, you can contribute it back to the library for others to use. The contribution process is structured around three quality tiers.

To submit your agent, open a new issue on the repository using the agent submission template at .github/ISSUE_TEMPLATE/agent-submission.md. Fill in the required fields, attach your files, and the maintainers will review your submission. The CONTRIBUTING.md file in the repository root contains the full submission guidelines, coding standards, and review criteria.

Community contributions are what keep the template count growing. The repository started with a core set of templates and has expanded to 162 through a combination of maintainer work and community submissions. The GitHub Discussions tab (specifically the "Share Your Agent" thread) is also an active space for sharing works-in-progress and getting feedback before formal submission.

Repository Stats and Community

The awesome-openclaw-agents repository has grown into one of the most active open-source AI agent resources on GitHub. Here are the current numbers.

Top Referral Sources

Traffic comes from a healthy mix of sources, which indicates organic discovery across multiple platforms:

The fact that ChatGPT is a top-5 referrer is noteworthy. It means people are asking ChatGPT about AI agent frameworks, and ChatGPT is recommending the OpenClaw repository. This is a strong signal of the project's visibility in the AI community.

Community Features

Beyond the templates themselves, the repository has several community features:

• - GitHub Discussions: Active threads for sharing agents, requesting templates, and asking questions. The "Share Your Agent" discussion is where community members showcase their builds.
• - Newsletter: A Google Form-based email newsletter for updates on new templates, featured agents, and repository milestones.
• - Agent Submission System: Structured issue templates for submitting new agents, with clear guidelines for each quality tier.
• - Category Search: The README includes a complete table of all categories with agent counts, making it easy to find what you need.

Get Started in 2 Minutes

Whether you are building your first agent or adding to an existing team, the fastest path is to start with a template. Here is the quickest workflow:

# 1. Install OpenClaw (if not already installed)
npm install -g openclaw

# 2. Clone the template repository
git clone https://github.com/mergisi/awesome-openclaw-agents.git
cd awesome-openclaw-agents

# 3. Browse categories and pick a template
ls agents/

# 4. Copy a template to your workspace
cp -r agents/software-engineering/full-stack-developer/ ~/my-agent/

# 5. Customize the SOUL.md to fit your needs
# (edit identity, rules, and workflow sections)

# 6. Register and run
cd ~/my-agent
openclaw agents add my-dev-agent ./
openclaw gateway start
openclaw agent --agent my-dev-agent --message "Review this pull request"

Or skip the terminal entirely and browse all 162 templates visually at crewclaw.com/agents. Filter by category, preview the SOUL.md, and deploy with a few clicks.

Related Guides

Frequently Asked Questions