🤖

Agents

Autonomous AI workflows and assistants that combine multiple skills to accomplish complex tasks.

3,426 total · 291 safe · 165 malicious

🤖

Paddleocr Doc Parsing

by bobholamovic

6

This skill helps you extract structured data, like tables and text, from complex documents such as PDFs and images using an external API.

review Marketing & Sales
0
🤖

Email Automation

by fly3094

6

This skill automates your email inbox by triaging, categorizing, and drafting replies to emails across Gmail, Outlook, and IMAP accounts.

review Communication
0
🤖

Openclaw Acp

by fcfsprojects

6

This skill allows agents to interact with the Agent Commerce Protocol (ACP) to discover and use other agents, create jobs, and manage agent services.

review CLI Utilities
0
🤖

Proactive Amcp

by fcavalcantirj

2

This skill aims to create a persistent digital identity and memory for an agent, storing its state on IPFS and allowing it to recover from failures.

malicious Uncategorized
0
🤖

Jungle Executor

by calwedbot-del

3

This agent enforces a set of 'Jungle Laws' and 'Warden Protocols' for decision-making, trading, and task execution, with features for rule validation, self-auditing, and automated correction or kill-switches.

malicious Productivity & Tasks
0
🤖

Use My Browser

by flyinghanger

3

This skill allows the AI to control your Chrome browser, navigate websites, fill out forms, and extract information by injecting code into web pages.

malicious Search & Research
0
🤖

Deai Marketplace

by fanglabgames

5

Connects an AI agent to the DeAI decentralized asset auction marketplace on the Base blockchain for buying, selling, and bidding on tokenized assets.

suspicious AI & LLMs
0
🤖

Proactive Agent Skill

by fangkelvin

6

This skill aims to make AI agents more proactive by implementing memory persistence, automated tasks, and self-improvement features.

review Productivity & Tasks
0
🤖

Company Search

by fan31415

4

This tool helps you research companies by gathering information on shareholders, legal representatives, funding, lawsuits, and more from various online sources.

suspicious Search & Research
0
🤖

Play Structs

by abstrct

4

This skill guides you through setting up and playing a game called Structs, which involves creating an account, building mining infrastructure, and refining in-game resources.

suspicious AI & LLMs
0
🤖

Evoclaw

by eyedark

4

This skill aims to help an AI agent develop and manage its own identity and beliefs over time, based on its experiences and interactions.

suspicious Coding Agents & IDEs
0
🤖

Mcp Server Setup Kit

by flynndavid

6

This skill guides users through setting up integrations between Claude AI and popular tools like GitHub, Notion, Slack, and Linear.

review Productivity & Tasks
0
🤖

Botlearnb Autodidact

by calvinxhk

6

This skill helps the AI agent learn and improve by automatically searching for and installing new tools or skills from a community repository to solve unsolved tasks.

review Productivity & Tasks
0
🤖

Persistent User Memory

by eyadhrif

6

This skill helps OpenClaw remember your preferences and past interactions to provide a more personalized and helpful experience.

review Security & Passwords
0
🤖

Linkdapi

by foontinz

6

This skill allows you to access LinkedIn profile and company data, search for jobs, and find people using the LinkdAPI service.

review Data & Analytics
0
🤖

Join Crabla

by abstrct

4

This skill helps an AI agent join a blockchain game guild called KC (Crabla) and manage in-game resources like mining and refining.

suspicious AI & LLMs
0
🤖

Agent Browser With Camoufox

by adastraabyssoque

3

This skill helps you install and configure a privacy-focused browser called Camoufox along with a browser automation tool called agent-browser, allowing it to work with Camoufox.

malicious Search & Research
0
🤖

Team Collab Skill

by aboutyao

6

This skill helps you set up a system where multiple AI agents can work together on tasks, remembering past interactions and sharing information.

review AI & LLMs
0
🤖

Super Brain

by aboutyao

3

This skill aims to give AI a persistent memory by storing user preferences, conversation history, and learning from interactions over time.

malicious AI & LLMs
0
🤖

Expanso Secrets Scan

by aronchick

3

This skill scans text or code for hardcoded secrets like API keys and passwords.

malicious Security & Passwords
0
🤖

Agentmail

by adboio

6

This skill allows AI agents to send and receive emails programmatically, manage dedicated email inboxes, and handle email-based workflows.

review Communication
0
🤖

Fibuki

by felixtosh

6

Connects to European bank accounts to manage transactions, match receipts, and categorize expenses through FiBuKI.com.

review Finance
0
🤖

Evolink Music

by evolinkai

7

This skill uses AI to generate music and songs based on your text descriptions, lyrics, and style preferences.

safe AI & LLMs
0
🤖

Solo Factory

by fortunto2

4

Installs a comprehensive toolkit for software development, including code intelligence and web search capabilities.

suspicious CLI Utilities
0

What Are OpenClaw Agents?

OpenClaw agents are autonomous AI workflows that combine multiple skills to accomplish complex tasks. Unlike individual skills which provide a single capability, agents orchestrate multiple skills together — planning steps, maintaining context through OpenClaw's Memory component, and executing multi-step workflows without human intervention.

Agents range from simple task automators to sophisticated assistants that can handle research, code generation, data analysis, and more. ClawGrid security-scores every agent so you can install with confidence.

Agents FAQ

What are OpenClaw agents?

Agents are autonomous AI workflows that combine multiple skills to accomplish complex tasks. They can plan, reason, and execute multi-step workflows without human intervention. Read our guide to skills, agents, and assistants.

How are agents different from skills?

Skills are individual capabilities (e.g., "read a file" or "query a database"). Agents combine multiple skills into autonomous workflows — they can plan, reason, and execute multi-step tasks. Think of skills as tools and agents as the workers who use them.

Can I build my own OpenClaw agent?

Yes — agents are defined using YAML configuration that specifies which skills to use, which LLM to run, and how to handle tasks. You can publish agents to ClawHub for others to discover. See our getting started guide.