Skip to content

EthanYolo01/Awesome-OpenClaw

BranchesTags

Repository files navigation

OpenClaw Ecosystem Map

🦞 Awesome OpenClaw

A carefully curated list of awesome OpenClaw resources — not everything, just the best.

Skills · Plugins · MCP · Tools · Deployments · Security · Research · Alternatives

Awesome PRs Welcome License: CC0-1.0 Last Updated Link Check Stars Contributors

Opinionated curation over exhaustive listing. Every entry was reviewed. Where other lists stop at a URL, we tell you why it matters and when to use it.

OpenClaw at a Glance (March 2026): ⭐ 247K GitHub Stars  ·  🔱 47.7K Forks  ·  🧩 13,729 ClawHub Skills  ·  🌍 10+ Channels

🌐 Language / 语言切换：English | 中文

📋 Table of Contents

🏁 Start Here — What is OpenClaw?

OpenClaw (formerly Clawdbot, briefly Moltbot) is an open-source, local-first autonomous AI agent framework. You run a persistent process called the Gateway on hardware you control — a Mac mini, a VPS, a Raspberry Pi — and it connects to messaging apps you already use. When you send a message, OpenClaw runs an agent turn powered by an LLM, invokes tools or skills, and responds — often taking real actions on your behalf.

Key design decisions that explain the hype:

  • Messaging app as UI — No new app to install. WhatsApp, Telegram, Discord, Slack, iMessage — whatever you already live in becomes your agent interface.
  • Local-first, self-hosted — Your data stays on your machine. No third-party server sees your context.
  • Skills/Plugins architecture — Capabilities are tiny markdown-declared plugins. Anyone can write one in under an hour.
  • Persistent memory — Flat-file Markdown memory (MEMORY.md, SOUL.md) persists across restarts without a vector database.
  • Model-agnostic — Works with Claude, GPT-5, Gemini, local Ollama models, and OpenRouter.

⏳ Timeline & Milestones

Understanding OpenClaw's history explains its current architecture and community dynamics.

Nov 2025  ──  Peter Steinberger (@steipete) releases "Clawdbot"
              TypeScript, Claude-only, WhatsApp + Telegram support
              Day 1: 5K stars. Week 1: 30K stars.

Dec 2025  ──  Anthropic trademark dispute → renamed "Moltbot"
              72 hours later renamed again to "OpenClaw"
              Community skill ecosystem begins forming spontaneously
              ClawHub launched as community skill registry

Jan 2026  ──  OpenClaw goes supernova: 68K → 120K stars in 2 weeks
              CVE-2026-25253 disclosed — WebSocket RCE vulnerability
              341 malicious skills discovered on ClawHub
              Alternative implementations start appearing:
              ZeroClaw (Rust), NanoClaw (Python), PicoClaw (Go)

Feb 2026  ──  Steinberger joins OpenAI
              Project transferred to OpenClaw Foundation (independent)
              ClawHub adds VirusTotal scanning for all submitted skills
              NemoClaw (NVIDIA, sandboxed) released
              Moltis (Rust, enterprise-grade) hits 1.0

Mar 2026  ──  247K stars, 47.7K forks
              OpenClaw Foundation announces governance model
              13,729 skills on ClawHub
              v1.9.x stable release series

🏠 Official Resources

Resource Description
openclaw/openclaw Main repository — TypeScript monorepo
openclaw/skills Official skills repository
clawhub.ai Official skills marketplace & registry
OpenClaw Blog Announcements, security advisories, feature deep-dives
OpenClaw Docs Official documentation
OpenClaw Changelog Release notes and version history
Latest Updates & Roadmap Official X account — feature previews, release news, community highlights

Monorepo package map:

Package Description
packages/core Core framework, provider interfaces, base classes
packages/gateway WebSocket control plane (port 18789), session management, channel routing
packages/agent Agent runtime with RPC mode, tool streaming, block streaming
packages/cli CLI for onboarding, gateway control, messaging
packages/sdk SDK for building custom tools and plugins
packages/ui WebChat interface + Control UI dashboard

⚡ Getting Started & Installation

Quick Install

# Requires Node.js 20+
npm install -g openclaw

# Interactive onboarding wizard
openclaw onboard

# Or specify your provider directly
openclaw onboard --auth-choice anthropic-api-key
openclaw onboard --auth-choice openai-api-key
openclaw onboard --auth-choice openrouter        # Recommended for cost optimization

Docker (Recommended for Production)

docker run -d \
  -p 127.0.0.1:18789:18789 \
  -v ~/openclaw-data:/data \
  -e ANTHROPIC_API_KEY=your_key \
  --restart unless-stopped \
  openclaw/openclaw:stable

⚠️ Critical: Bind to 127.0.0.1, not 0.0.0.0. See Security for why this matters.

Release Channels

Channel Command Use Case
stable npm i -g openclaw All users — recommended
beta npm i -g openclaw@beta Early adopters, testing new skills
dev npm i -g openclaw@dev Core contributors only — may be broken

Install Guides

🧠 Core Architecture

Five concepts unlock the entire ecosystem:

1. The Gateway

A WebSocket server (port 18789) that is the single orchestration point. It normalizes heterogeneous messaging protocols — WhatsApp via Baileys, Telegram via grammY, Slack Events API, Discord gateway — into a canonical internal message format. No intelligence lives here. It is a pure traffic controller.

2. Skills vs Plugins vs MCP

Skills Plugins MCP Integrations
Format SKILL.md markdown TypeScript npm package Separate process, any language
Complexity Zero-code Full code + lifecycle hooks Full code + separate runtime
Distribution ClawHub registry npm / GitHub Any URL or local
Isolation None (in-process) Application-level Process-level
Best for Quick capabilities, integrations Deep system access, custom providers External service integrations
Example Gmail skill, Notion skill ClawRouter, SecureClaw Playwright MCP, GitHub MCP

3. Memory Architecture

OpenClaw uses filesystem-as-truth memory — no vector database required by default.

File Purpose Mutable by Agent?
MEMORY.md Long-term distilled facts ✅ Yes
SOUL.md Core identity and values ✅ Yes (unless locked)
AGENTS.md Immutable operating instructions ❌ No — agent cannot touch this
memory/YYYY-MM-DD.md Daily episodic logs ✅ Auto-appended

4. Context Engineering

When approaching token limits, OpenClaw triggers session compaction — summarizing oldest turns while preserving tool call-result pairs. Sessions reset at configurable boundaries (default: 4 AM UTC). Result: effectively unlimited conversation length.

5. Channel Isolation

Each connected messaging platform is a channel with its own context namespace. Skills can be scoped per channel using channels: in the skill manifest. Run 10+ channels simultaneously with per-channel permission sets.

🤖 LLM Model Selection Guide

This is one of the most common questions new users ask, and no other resource covers it well. Here is the community consensus as of March 2026.

Model Comparison at a Glance

Model Provider Strength Weakness Best OpenClaw Use Case
Claude Sonnet 4 Anthropic Best tool-calling reliability; nuanced instruction following Higher cost than Gemini Default recommendation; daily driver for most users
Claude Haiku 4 Anthropic Very fast, cheap Less capable for complex multi-step tasks High-volume triage, simple automations, notifications
GPT-5 OpenAI Strong reasoning, very capable coding Cost; sometimes over-verbose in tool calls Complex coding tasks, analytical workflows
Gemini 2.5 Pro Google Longest context window (1M tokens); best price at scale Tool call formatting occasionally inconsistent Document-heavy workflows, research, long sessions
Gemini 2.5 Flash Google Fastest response, cheapest capable model Less reliable on complex tool chains Real-time automations, cost-sensitive deployments
Llama 4 Scout (Ollama) Meta / Local Free, fully private, no API key needed Needs capable hardware (16GB+ VRAM recommended) Air-gapped environments, privacy-critical data
Qwen3-32B (Ollama) Alibaba / Local Strong multilingual; excellent for non-English workloads Hardware requirements Multilingual deployments
OpenRouter Multi Automatic routing, unified billing, fallback logic Adds latency, another dependency Teams wanting flexibility; ClawRouter power users

Recommendation by Use Case

Personal productivity agent (email, calendar, tasks): → Claude Sonnet 4 as primary. Claude Haiku 4 for simple reminders and quick lookups. Route with ClawRouter.

Software development assistant: → GPT-5 or Claude Sonnet 4. Both excel at code. GPT-5 slightly stronger on large refactors; Sonnet 4 more reliable on multi-step tool chains.

Research and document processing: → Gemini 2.5 Pro for its 1M token context window. Process entire codebases or large PDFs without chunking.

Privacy-first / air-gapped deployment: → Llama 4 Scout (16GB VRAM) or Qwen3-14B (8GB VRAM) via Ollama. No data leaves your machine.

Cost-optimized high-volume: → Gemini 2.5 Flash or Claude Haiku 4. Use ClawRouter to auto-route cheap tasks to these and expensive tasks to Sonnet/GPT-5.

Configuring Multiple Models

# Configure a primary and fallback model
openclaw config set model.primary "Claude Sonnet 4.6"
openclaw config set model.fallback "gemini-2.5-flash"
openclaw config set model.coding "gpt-5"

# Or use OpenRouter for automatic routing
openclaw config set provider "openrouter"
openclaw config set model.primary "openrouter/auto"

Local Model Setup with Ollama

# Install Ollama
curl -fsSL https://ollama.ai/install.sh | sh

# Pull a recommended model
ollama pull llama4:scout        # 16GB VRAM
ollama pull qwen3:14b           # 8GB VRAM
ollama pull mistral-nemo:12b    # 6GB VRAM (minimum viable)

# Point OpenClaw at local Ollama
openclaw config set provider "ollama"
openclaw config set model.primary "llama4:scout"

💰 Cost Estimation Guide

One of the top reasons people hesitate to try OpenClaw. Here are real numbers.

API Cost by Provider (March 2026 Pricing)

Model Input (per 1M tokens) Output (per 1M tokens) Typical turn cost
Claude Sonnet 4 $3.00 $15.00 ~$0.005–0.02
Claude Haiku 4 $0.80 $4.00 ~$0.001–0.005
GPT-5 $10.00 $30.00 ~$0.01–0.05
Gemini 2.5 Pro $1.25 $5.00 ~$0.003–0.015
Gemini 2.5 Flash $0.15 $0.60 ~$0.0003–0.002
Llama 4 / Ollama $0 $0 Hardware cost only

"Typical turn cost" = one user message + agent response, including tool calls. Complex multi-step tasks cost more.

Monthly Cost Estimates by Usage Profile

Profile Description Est. Monthly API Cost
Light ~50 messages/day, mostly simple tasks $3–8/month
Moderate ~200 messages/day, mix of simple and complex $15–40/month
Heavy ~500 messages/day, complex workflows with many tool calls $50–150/month
Team (5 people) Shared instance, ~1000 messages/day $100–300/month
Enterprise Custom; ClawTeam managed billing Contact vendors

Actual costs vary dramatically based on model choice, skill complexity, and conversation length. The cost-tracker skill gives you real per-conversation data.

Cost Optimization Strategies

1. Use ClawRouter for intelligent routing (40–60% savings)

openclaw skill install BlockRunAI/clawrouter
# Routes simple tasks (reminders, lookups) to Haiku/Flash
# Routes complex tasks (code, research) to Sonnet/GPT-5
# Users report 40-60% cost reduction

2. Context window hygiene

# Enable aggressive session compaction
openclaw config set context.compaction_threshold 0.7
openclaw config set context.reset_schedule "0 4 * * *"  # 4am UTC daily

3. Scope skills narrowly Skills with broad triggers (use this when the user asks anything about...) consume more tokens evaluating whether to activate. Narrow trigger conditions = fewer wasted tokens.

4. Use local models for bulk tasks Run Ollama locally for high-volume, lower-complexity tasks (email sorting, routine summaries). Zero marginal API cost.

Total Cost of Ownership: Self-Hosted vs Managed

Self-Hosted VPS Managed (ClawHost) Self-Hosted Homelab
Infrastructure $5–18/month (Hetzner/DO) $10–30/month ~$0/month (existing hardware)
Setup time 30–120 min 5–15 min 2–4 hours
Maintenance 1–2 hrs/month None 1–2 hrs/month
Privacy High Medium Maximum
Reliability High (99.9% SLA typical) High Depends on hardware
Best for Most users Beginners, busy professionals Power users, homelabbers

🔧 Skills & Plugins

Official Skill Registry (ClawHub)

clawhub.ai — The official marketplace. 13,729 community-built skills as of March 2026.

openclaw skill install steipete/slack
openclaw skill install community/playwright-mcp
openclaw skill list
openclaw skill remove steipete/slack

⚠️ ClawHub is open-submission. Not all skills are vetted. Always review source before installing. See the Security section.

Curated "Must-Have" Skills

🛡️ Install This First:

Skill What it does Why essential
secureclaw Audits your install for CVEs, misconfigurations, prompt injection risks Install before anything else

Productivity Core:

Skill What it does Stars / Installs
playwright-mcp Full browser automation — browse, click, scrape, fill forms Most-installed skill
agentmail AI email triage, auto-replies, LLM-driven inbox management 30–50% inbox time reduction (reported)
notion-direct Bidirectional Notion sync Best knowledge management integration
obsidian-direct Reads/writes Obsidian vault directly For Markdown-native PKM users
google-calendar Full calendar management, natural language "Book a meeting Tuesday afternoon" just works
linear Create, update, triage Linear issues via chat Essential for engineering teams
github-mcp Full GitHub integration via MCP PR reviews, issue management, code search
composio Connect to 250+ apps via managed auth Best for broad SaaS coverage

Development:

Skill What it does
claude-code Runs Claude Code as a sub-agent for coding tasks
docker-manager Manage containers, images, and compose stacks
cost-tracker Real-time LLM API cost monitoring and budget alerts
cron-backup Scheduled backups with version tracking

Research & Data:

Skill What it does
web-research Multi-source web research with citation tracking
arxiv-search Search and summarize arXiv papers
perplexity-search Deep web search via Perplexity API
supermemory External memory layer backed by Supermemory.ai

Skills by Category

For the full categorized index of 5,400+ vetted skills: ➡️ VoltAgent/awesome-openclaw-skills — 1M+ monthly views, the #1 community skills resource.

Building Your Own Skills

A skill is a SKILL.md file. Minimum viable skill:

# my-skill

Use this skill when the user asks to [TRIGGER CONDITION].

## Setup
- Requires: API_KEY environment variable

## Tools

### do_thing
Does the thing.
Parameters:
- query (string): What to do

## Instructions
When triggered, call do_thing with the user's request.
Return the result in a friendly format.

Skill authoring resources:

🤖 MCP Integrations

MCP (Model Context Protocol, originally by Anthropic) is the dominant architecture for new professional-grade OpenClaw skills. MCP servers run as independent processes, offering better isolation and language-agnosticism.

npm install -g mcporter
openclaw mcp add --url https://mcp.github.com/sse --name github
openclaw mcp add --local /path/to/my-mcp-server --name local-tools

Notable MCP Servers

Server Category Notes
playwright-mcp Browser Most-used MCP in OpenClaw; full browser automation
github-mcp DevOps Official GitHub MCP
filesystem-mcp System Read/write files with explicit path control
memory-mcp Memory Knowledge graph memory
brave-search-mcp Search Web search via Brave API
fetch-mcp HTTP Web fetching and scraping
postgres-mcp Database Read-only PostgreSQL access
slack-mcp Comms Official Slack MCP
gmail-mcp Email Gmail via MCP

Browse 13,000+ MCP servers at mcp.so or the official MCP server directory.

🖥️ UIs & Companion Apps

Web UIs

Tool Description Key Feature
openclaw-studio Full-featured web dashboard Skill management, memory viewer, cost analytics
vibeclaw Browser-native interface — no local install Test skills before production deploy
openclaw-ui Official WebChat + Control UI Ships with core
ClawDash OpenClaw agent interface Advanced Next.js 15 UI boilerplate with 50+ components, modular layout and professional dashboard design
OpenClaw Directory Community skill/tool catalog Searchable, community-rated

Desktop Apps

Tool Platform Description
OpenClaw for macOS macOS Menu bar companion app; manages/connects to local Gateway and exposes macOS capabilities as nodes to the agent
OpenClaw Manager macOS / Windows Desktop chat interface with visual file management
OpenClaw Windows Hub Windows Windows companion suite for OpenClaw (AI personal assistant)

Mobile

Tool Platform Description
OpenClaw iOS App iOS Internal preview — not yet publicly released. Connects to Gateway via WebSocket (LAN or tailnet); exposes node capabilities: canvas, screenshot, camera capture, location, call mode, voice wake. Receives node.invoke commands and reports node status events
OpenClaw Android Node Android Official Android platform docs; community APK build at openclaw-android-node-apk

🌐 Platform Channels & Messaging

Platform Status Notes
Telegram ✅ Native Best-supported; works on all implementations
WhatsApp ✅ Native (Baileys) Most popular for personal use; requires QR scan
Discord ✅ Native Guild + DM; slash commands available
Slack ✅ Native Full Events API + Block Kit
iMessage ✅ macOS only Needs BlueBubbles bridge on Linux/Windows
Signal ✅ Via signal-cli Best privacy option
Google Chat ✅ Native Workspace integration
Microsoft Teams ✅ Beta Bot Framework integration
SMS ✅ Via Twilio Pay-per-message
Email ✅ Via skills IMAP/SMTP or AgentMail plugin
Voice 🧪 Experimental OpenAI Realtime API, early-stage
Web widget ✅ Via openclaw-ui Embeddable for websites

Multi-channel tips:

  • Each channel has its own conversation namespace — skills can be scoped per channel
  • Dedicate a cheap Telegram bot to low-priority tasks; keep WhatsApp for high-priority work
  • Channel-level permission sets: restrict dangerous skills to trusted channels only

☁️ Deployment & Infrastructure

Managed Hosting

Service Starting Price Best For
ClawHost $25–$350/month Choose from 45+ provider servers to match your needs
Tinkerclaw ~$49/month Full onboarding support + OpenClaw Manager
OpenClaw Launch ~$20/month Simple managed + pre-configured skills
DigitalOcean 1-Click $12/month Self-hosted but fast; full root access

Self-Hosted: Docker Compose

version: '3.8'
services:
  openclaw:
    image: openclaw/openclaw:stable
    restart: unless-stopped
    ports:
      - "127.0.0.1:18789:18789"   # NEVER bind to 0.0.0.0 directly
    volumes:
      - ./data:/data
      - ./skills:/skills
    environment:
      - ANTHROPIC_API_KEY=${ANTHROPIC_API_KEY}
      - OPENCLAW_MEMORY_DIR=/data/memory
      - OPENCLAW_LOG_LEVEL=warn
    user: "1000:1000"
    read_only: true
    tmpfs: ["/tmp"]
    security_opt:
      - no-new-privileges:true
    healthcheck:
      test: ["CMD", "curl", "-f", "http://localhost:18789/health"]
      interval: 30s
      timeout: 10s
      retries: 3

Self-Hosted: NixOS

services.openclaw = {
  enable = true;
  package = pkgs.openclaw;
  settings = { port = 18789; memoryDir = "/var/lib/openclaw/memory"; };
  environmentFile = "/run/secrets/openclaw-env";
};

See nix-openclaw for the community NixOS module.

VPS Price Comparison

Provider Tier Cost Notes
Hetzner CX22 (2GB) €4.15/mo Best price/performance in EU
DigitalOcean Basic 2GB $18/mo Most tutorials target this; 1-click available
Vultr Regular 2GB $12/mo Good global coverage
Linode/Akamai Nanode 1GB $5/mo Tight on RAM; upgrade if adding many skills
Oracle Cloud Always Free 1GB $0 Free; performance varies

Homelab & Edge

Resource Notes
Raspberry Pi Guide Pi 4 4GB+ required; Pi 5 preferred
Mac Mini Setup Recommended: silent, efficient, reliable
Ansible Playbook Automated install with Tailscale VPN, UFW firewall and Docker isolation
OpenClaw on Proxmox Community tutorial: complete Proxmox VE deployment guide

🏢 Enterprise Deployment

For teams and organizations with compliance, multi-tenancy, and governance requirements.

Authentication & Access Control

Single Sign-On (SSO):

# SAML 2.0 via Nginx auth_request
# OpenClaw itself is auth-agnostic; put auth in your reverse proxy layer
# Community guide: https://docs.openclaw.ai/enterprise/sso
Tool Description
agent-access-control skill Tiered access control — restrict which users trigger which actions

Multi-Tenancy Patterns

OpenClaw does not have native multi-tenancy. Enterprise teams use one of three patterns:

Pattern A — One instance per team/user (Recommended)

User A → Gateway A (port 18789) → Memory A
User B → Gateway B (port 18790) → Memory B

Complete isolation. Simple. Works with Ansible or Docker Compose per user.

Pattern B — Shared instance, per-channel isolation One Gateway, multiple channels, channel-scoped skills and permissions. Lower cost; shared memory is a risk.

Pattern C — ClawTeam managed multi-tenancy ClawTeam handles auth, routing, and billing. Best for teams >10 people who don't want to manage infra.

Compliance & Auditing

Tool Description Standard
agent-audit-trail Tamper-evident hash-chained action logs SOC 2 Type II compatible
SecureClaw 55-point security audit; OWASP ASI + MITRE ATLAS mappings OWASP ASI Top 10
NemoClaw NVIDIA-sandboxed build; GPU-accelerated, fully isolated Air-gap ready

Data Residency

By default, your data goes to your chosen LLM provider's API. For organizations with data residency requirements:

  • EU: Use Anthropic EU endpoints or Gemini EU-hosted endpoints; deploy OpenClaw on Hetzner Germany
  • Air-gap: Run Ollama locally — zero data leaves your network. See Ollama × OpenClaw Integration Guide
  • On-premise: NemoClaw for fully isolated deployments with local GPU inference

Disaster Recovery

# Backup your entire OpenClaw state
tar -czf openclaw-backup-$(date +%Y%m%d).tar.gz \
  ~/openclaw-data/memory \
  ~/openclaw-data/skills \
  ~/.openclaw/config.json

# Restore
tar -xzf openclaw-backup-20260315.tar.gz -C ~/openclaw-data/

The cron-backup skill automates this on a schedule and optionally uploads to S3, Backblaze B2, or any S3-compatible provider.

🧠 Memory & Knowledge Systems

Built-in Memory Files

File Purpose Mutable by Agent
MEMORY.md Long-term distilled facts and preferences ✅
SOUL.md Core identity, values, operating instructions ✅ (lock recommended)
AGENTS.md Immutable rules — agent cannot override ❌
memory/YYYY-MM-DD.md Daily episodic logs ✅ Auto-appended

External Memory Integrations

Tool Description Best For
Supermemory Managed external memory with semantic search Teams needing shared memory across instances
Graphiti Temporal knowledge graph memory Complex relational knowledge over time
MemOS Hierarchical memory OS for LLM agents Research-grade memory architecture
mem0 Self-improving personalized memory Personalization-heavy workflows

Memory Best Practices

  1. Keep SOUL.md short and opinionated. 200–300 words. Specific beats generic.
  2. Use AGENTS.md for non-negotiable rules. Security policies, content restrictions, tool limits — the agent cannot override it.
  3. Prune MEMORY.md monthly. Old facts accumulate. Review and remove stale entries.
  4. Separate work and personal contexts using channel-scoped memory with different channel configurations.
  5. Never put credentials in memory files. Use environment variables or the secrets manager.

🔒 Security

This section exists because no other Awesome OpenClaw list treats security seriously enough. OpenClaw runs with broad system access. Security is not optional.

Known CVEs & Incidents

CVE / Incident Severity Description Mitigation
CVE-2026-25253 🔴 Critical WebSocket RCE — unauthenticated remote code execution if gateway exposed to internet Update to v1.8.4+. Never expose port 18789 directly.
ClawHub Malicious Skills (Jan 2026) 🔴 High 341 malicious skills; some exfiltrated data via Base64-chunked requests to unknown endpoints. 26% of popular skills had at least one high-severity vector (Cisco research) Use SecureClaw + skill-guard before installing any skill
41K+ Exposed Instances (Feb 2026) 🔴 High 41,600 OpenClaw gateways publicly exposed; 2,100+ leaking API keys Audit firewall rules. Use the SecureClaw network audit.
Prompt Injection via Skill Descriptions 🟠 Medium Malicious skills embed hidden instructions overriding user intent SecureClaw behavioral rules; review skill source
API Key in Memory Logs 🟡 Low–Medium Agent wrote API keys to daily logs when processing config messages Fixed in v1.9.1; audit old memory/ logs

Essential Security Tools

SecureClaw (Install First)

SecureClaw by Adversa AI:

  • 55 audit checks across the full attack surface
  • 15 behavioral rules (~1,150 tokens) against prompt injection
  • Maps to OWASP ASI Top 10 and MITRE ATLAS
openclaw plugin add secureclaw
openclaw skill install adversa-ai/secureclaw
openclaw secureclaw audit --full
openclaw secureclaw harden

Other Security Tools

Tool Description
skill-guard Scan ClawHub skills for security vulnerabilities before installing — detects prompt injection, malware payloads, hardcoded keys and other threats
agent-audit-trail skill Tamper-evident hash-chained action logging
agent-access-control skill Tiered access control per user/channel

Secure Deployment: Nginx Reverse Proxy

Never expose port 18789 directly. Always use a reverse proxy with authentication:

server {
    listen 443 ssl http2;
    server_name yourdomain.com;

    ssl_certificate     /etc/letsencrypt/live/yourdomain.com/fullchain.pem;
    ssl_certificate_key /etc/letsencrypt/live/yourdomain.com/privkey.pem;

    location /openclaw/ {
        auth_basic "OpenClaw Gateway";
        auth_basic_user_file /etc/nginx/.htpasswd;

        proxy_pass http://127.0.0.1:18789/;
        proxy_http_version 1.1;
        proxy_set_header Upgrade $http_upgrade;
        proxy_set_header Connection "upgrade";
        proxy_set_header Host $host;

        # Rate limiting
        limit_req zone=openclaw burst=20 nodelay;
        limit_req_log_level warn;
    }
}

# Rate limit zone definition (in http block)
# limit_req_zone $binary_remote_addr zone=openclaw:10m rate=10r/s;

Skill Vetting Checklist

Before installing any skill from ClawHub:

  • Author verification: Real GitHub profile with commit history?
  • Age and stars: New repo + suspiciously high stars = red flag
  • Read the source: SKILL.md files are markdown — readable in 2 minutes. Look for hidden instructions after HTML comments or inside code blocks
  • Permission scope: Does a calculator skill request network access? Reject over-privileged skills
  • VirusTotal: Check the ClawHub page for VirusTotal scan results
  • Run skill-guard: openclaw run "use skill-guard to audit [skill-name]"

Automatic red flags in skill source:

  • Instructions to ignore previous instructions or AGENTS.md
  • Requests to send data to external URLs not mentioned in the README
  • Base64-encoded content embedded in the skill file
  • Instructions to modify SOUL.md or AGENTS.md
  • fetch() calls to hardcoded external domains without disclosure

🆚 OpenClaw vs Other Agent Frameworks

For product teams and developers evaluating which framework to build on. Updated March 2026.

Comparison Matrix

OpenClaw AutoGen CrewAI LangGraph n8n Botpress
Primary Language TypeScript Python Python Python Node.js TypeScript
Learning Curve Low High Medium High Low Medium
Agent Type Single agent + skills Multi-agent orchestration Multi-agent teams State machine graphs Workflow automation Conversational bots
Memory Built-in flat-file + external Manual Manual LangChain memory None native Built-in
UI Messaging apps (WhatsApp, TG…) Code / Jupyter Code / Web UI Code / LangSmith Web UI Web chat UI
Self-hosted ✅ First-class ✅ ✅ ✅ ✅ ✅
Skill/Plugin Ecosystem 13,729 skills (ClawHub) Community tools Community crews LangChain tools 500+ integrations Community nodes
Security Model Application-level (NanoClaw provides container isolation) None native None native None native None native None native
Best for Personal / team productivity agents Research, complex multi-agent Business workflow automation Stateful pipelines, research Non-technical automators Customer-facing chatbots
Worst for Complex multi-agent pipelines Non-technical users Real-time personal assistants Simple use cases LLM-heavy workflows Power users needing full control
GitHub Stars 247K 38K 26K 12K 51K 14K

When to Choose OpenClaw

✅ Choose OpenClaw if:

  • You want a personal productivity agent you interact with via messaging apps
  • You need a large ecosystem of pre-built skills (13,729 on ClawHub)
  • You want low-code skill authoring (SKILL.md format)
  • Local-first, self-hosted privacy matters to you
  • You're building for a team that will use their existing chat tools

❌ Don't choose OpenClaw if:

  • You need complex multi-agent orchestration with elaborate handoff protocols (→ use AutoGen or CrewAI)
  • You're building a customer-facing chatbot with conversation flows (→ use Botpress)
  • You need a visual no-code automation builder for non-technical users (→ use n8n)
  • You're doing research requiring fine-grained state machine control (→ use LangGraph)

OpenClaw vs AutoGen

The most common comparison. Key difference: AutoGen is designed for multi-agent systems where agents collaborate in code. OpenClaw is designed for single-agent productivity with a rich skill ecosystem. If you want an AI assistant that takes actions on your behalf via chat, OpenClaw wins. If you want agents collaborating to solve complex research problems, AutoGen wins.

OpenClaw vs n8n

n8n is a visual workflow automation tool for non-technical users. OpenClaw is a conversational agent for power users. They're complementary: many users run n8n for scheduled workflows and OpenClaw for on-demand conversational tasks. The n8n-webhook skill bridges the two.

🌍 Ecosystem & Agent Platforms

Agent Social & Collaboration

Platform Description
Moltbook Agent-native social network — agents post, follow, and interact
Clawk 𝕏/Twitter for AI Agents — the place to discover what agents are up to, up to 400 characters at a time
AgentDo Task marketplace — post tasks for agents or have your agent pick up work

Agent-to-Agent (A2A)

Resource Description
A2A Protocol Spec Open agent interoperability protocol, initiated by Google
ACP — Agent Communication Protocol Open agent interoperability protocol designed to address the growing challenge of connecting AI agents, applications and more
agent-team-orchestration skill Multi-agent teams with roles, tasks, handoffs, review
agentgate skill API gateway with human-in-the-loop write approval

🛠️ Developer Tooling

Deployment & Management

Tool Description
openclaw-ansible Official automated install with Tailscale VPN, UFW firewall and Docker security hardening

Development & Testing

Tool Description
Plugin SDK Documentation Plugin SDK ships as sub-path exports within the main package (openclaw/plugin-sdk/core, /telegram, /discord, etc.) — no separate package
Vibeclaw Browser-based sandbox — test without touching production
ClawHub CLI Official publishing tool: clawhub publish, clawhub inspect, versioning, soft-delete/restore

Monitoring & Observability

Tool Description
openclaw-dashboard Secure real-time monitoring dashboard with authentication, TOTP MFA, cost tracking, live data streaming and memory browser
cost-tracker skill Per-conversation API cost tracking with budget alerts

Cost Optimization

Tool Description
ClawRouter Agent-native LLM router supporting 41+ models, sub-1ms routing latency, USDC payments via x402 on Base and Solana
Model Router Routes model requests based on configured models, costs and task complexity — general/low-complexity requests to the cheapest available model, higher-complexity to stronger models

🏗️ Alternative Implementations

Decision Matrix

OpenClaw NanoClaw ZeroClaw Moltis PicoClaw Nanobot
Language TypeScript Python Rust Rust Go Python
RAM ~1.52 GB ~100 MB ~7.8 MB ~120 MB <10 MB ~110 MB
Binary 28 MB+ N/A 3.4 MB ~12 MB <10 MB N/A
Startup ~6 sec ~2 sec <10 ms ~1.5 sec ~1 sec ~3 sec
Isolation Application Container Multi-layer Process None None
Skills Ecosystem 13,729 (ClawHub) Via skills MCP-based Limited Limited MCP-based
Best For Max features Security-first Edge/IoT Enterprise Embedded Learning

Implementations

OpenClaw — Reference implementation. Choose when: you want the largest skill ecosystem and accept the resource tradeoffs as the cost of maturity.

NanoClaw — Python, mandatory container isolation, ~700 lines (auditable in an afternoon). Choose when: security and auditability matter more than skill count. Regulated environments.

ZeroClaw — Rust, 3.4MB binary, <10ms startup, 7.8MB RAM (194× smaller than OC), direct OpenClaw config import, 22+ providers. Choose when: resource efficiency matters, or you want a migration path from OpenClaw.

Moltis — A trusted Rust-native Claw. One binary — sandboxed, secure, auditable. Built-in voice, memory, MCP tools and multi-channel access. Choose when: your team needs a trustworthy enterprise-grade Rust implementation.

PicoClaw — Go, <10MB RAM, 1-second startup, targets $10 RISC-V hardware. Choose when: IoT, home automation, embedded Linux, cheapest possible hardware.

Nanobot — Python, ~4,000 lines, maximum readability, MCP-only tools. Choose when: learning agent internals, prototyping architectures, academic research.

NullClaw — Zig, 1MB RAM, 678KB binary. Choose when: absolute minimal footprint, microcontroller-adjacent.

Project Language Key Idea
ClawGo Go Headless, voice + text, embedded Linux
MimiClaw C ESP32 microcontroller version
Carapace Various Governance/compliance first

📖 Tutorials, Blogs & Articles

Getting Started

Intermediate & Advanced

Security

Alternatives & Ecosystem

Learning Paths

🎥 Videos & Talks

Resource Type Description
Lex Fridman #491: OpenClaw — The AI Agent Breaking the Internet Interview (3h15m) Creator Steinberger in deep conversation with Lex Fridman: origin story, architecture, Moltbook controversy, security risks, the future of autonomous agents. Best first video for understanding OpenClaw
OpenClaw Creator: Full Real-World Demo Demo interview Peter Yang (Creator Economy) talks to Steinberger: automated flight check-in, smart home control, counterintuitive AI coding workflows (no Plan Mode, no MCP). 2.71M views
"I Ship Code I Haven't Read" — OpenClaw Creator Interview Podcast / video Pragmatic Engineer (Gergely Orosz) interviews Steinberger in London just before the viral explosion — deep dive into AI-assisted development workflows. Available on YouTube and Spotify
OpenClaw Creator × OpenAI Developer Experience Lead Podcast OpenAI Builders Unscripted episode 1: Steinberger and Romain Huet on the early journey and advice for AI builders
50 Days of Real Workflows: 20 Prompt Templates Companion resource GitHub Gist with 20 real, copy-paste-ready prompts: morning briefings, backups, YouTube analytics, multilingual onboarding assistants and more

📚 Research & Academic Resources

Key Papers

Paper Relevance
Defensible Design for OpenClaw (2026) Comprehensive security framework; maps attack surface to OWASP ASI
ToolSword: Safety Issues in Tool Learning (2024) Foundational work on tool-invocation safety — directly applicable to skills
AgentBench: Evaluating LLMs as Agents (2023) Benchmark framework; used to evaluate OpenClaw variants
Prompt Injection Attacks (2023) Defines the prompt injection threat model affecting skill security
ReAct: Reasoning + Acting (2022) The ReAct pattern that OpenClaw's agent loop is based on

Academic Projects

  • OWASP Agentic Security Initiative — OpenClaw used as reference implementation for ASI Top 10
  • MITRE ATLAS — Globally accessible, continuously updated knowledge base of adversarial tactics and techniques against AI systems, based on real-world attack observations from AI red teams and security groups; SecureClaw includes formal MITRE ATLAS mappings
  • Nanobot — Best platform for studying agent internals; easy to instrument

💼 Use Cases & Showcases

By Domain

Personal Productivity:

  • Email → zero: Route all email through OpenClaw; auto-triage, draft for approval, auto-send routine responses
  • Calendar management: Natural language scheduling that actually understands context
  • Personal CRM: Remember context from all your relationships; briefed before meetings automatically

Software Development:

  • PR review assistant: Daily summaries of open PRs across all your repos
  • Incident responder: Monitors logs, pages on-call, drafts incident reports
  • Dependency auditor: Weekly scan for outdated or vulnerable packages

Business Operations:

  • Meeting notetaker: Record, transcribe, summarize, distribute
  • CRM intelligence: Monitor email, update HubSpot/Salesforce automatically
  • Report generator: Weekly metrics drafted and sent every Monday morning

Research & Data:

  • Literature monitor: Daily digest of new arXiv papers matching your interests
  • Competitive intelligence: Monitor competitors' GitHub, blog posts, job postings
  • Data pipeline: Scrape → clean → analyze → summarize on a schedule

Showcase Collections:

❓ FAQ & Troubleshooting

Installation & Setup

Q: What are the minimum hardware requirements? A: Node.js 20+ for OpenClaw core. Minimum 1GB RAM (2GB recommended for comfortable operation with several skills). Any modern 64-bit OS. For local models via Ollama: 8GB VRAM minimum; 16GB recommended.

Q: What's the difference between openclaw onboard and openclaw start? A: onboard is a one-time interactive wizard that sets up your config, connects your first channel, and installs essential skills. start just starts the gateway with existing config. Always run onboard first.

Q: Do I need a paid API key to get started? A: No. You can run fully locally using Ollama + an open-source model (Llama 4, Qwen3). The experience is less capable but costs nothing. Most users start with a Claude or OpenAI API key for the best out-of-the-box quality.

Gateway & Connection Issues

Q: Gateway starts but I can't connect from my messaging app. A: Check these in order:

  1. Confirm port 18789 is not blocked: curl http://localhost:18789/health
  2. If using Docker, ensure you're binding correctly: 127.0.0.1:18789:18789 (not 0.0.0.0)
  3. If accessing remotely, confirm your reverse proxy is correctly configured
  4. Check logs: openclaw logs --tail 50

Q: WhatsApp keeps disconnecting every few hours. A: This is a Baileys (WhatsApp Web library) limitation. Solutions:

  • Keep the QR-scanned device active and connected to the internet
  • Configure auto-reconnect rules in AGENTS.md, or use a deployment script with watchdog support
  • Consider switching to Telegram as your primary channel — it's more stable for server deployments

Q: My Telegram bot stopped responding. A: 99% of the time this is a bot token issue or a conflicting webhook. Check:

openclaw channel status telegram
openclaw channel reconnect telegram

Performance & Memory

Q: OpenClaw is using 2GB+ of RAM. Is that normal? A: With the full skill ecosystem loaded, yes. Mitigation strategies:

  • Use openclaw skill list --active to audit which skills are loaded
  • Uninstall skills you don't use: openclaw skill remove [name]
  • Consider ZeroClaw (7.8MB RAM) if resource efficiency is critical

Q: Responses are getting slower over long conversations. A: Context window is filling up. Solutions:

  • Enable aggressive compaction: openclaw config set context.compaction_threshold 0.7
  • Schedule daily resets: openclaw config set context.reset_schedule "0 4 * * *"
  • Configure health checks and auto-restart in your docker-compose.yml

Q: Memory is growing unboundedly. MEMORY.md is now 50,000 words. A: The agent will write memories indefinitely if not pruned. Best practices:

  • Set a monthly calendar reminder to review and prune MEMORY.md
  • Use the memory-manager skill to auto-summarize and compress old entries
  • Hard limit: Add to AGENTS.md: "Keep MEMORY.md under 10,000 words. Summarize and compress when approaching that limit."

Skills & Security

Q: I installed a skill and now my agent is behaving strangely. A: Likely prompt injection via a malicious or poorly written skill. Steps:

  1. Disable the skill: openclaw skill disable [name]
  2. Review its source code for hidden instructions
  3. Run openclaw secureclaw audit --full
  4. Report to ClawHub if it's genuinely malicious

Q: How do I prevent a skill from accessing the internet? A: Add to AGENTS.md:

[skill-name] must never make external network requests except to [approved-domain].

For stronger enforcement, use agent-access-control skill or switch to NanoClaw which offers container-level isolation.

Q: My API key was accidentally written to a memory log. What do I do?

  1. Immediately rotate the key at your provider's console
  2. Delete the affected log: rm ~/openclaw-data/memory/[date].md
  3. Run openclaw secureclaw audit to check for other exposures
  4. Add to AGENTS.md: "Never write, log, or repeat API keys, tokens, or credentials under any circumstances."

Costs & Models

Q: My API costs are much higher than expected. A: Common causes:

  1. Skills with very broad trigger conditions → audit with cost-tracker skill
  2. No session compaction → conversations accumulate context indefinitely
  3. Wrong model for the task → use ClawRouter to route cheap tasks to cheap models
  4. Context window leaks from a poorly written skill loading large documents

Q: Can I set a monthly spending cap? A: Not natively in OpenClaw core, but:

  • The cost-tracker skill sends alerts when you hit thresholds
  • Set hard limits directly at your API provider (Anthropic, OpenAI all support this)
  • ClawRouter can be configured to stop routing to expensive models after a cost limit

Updates & Migrations

Q: How do I safely update OpenClaw?

# Backup first
tar -czf ~/openclaw-backup-$(date +%Y%m%d).tar.gz ~/openclaw-data/

# Update
npm update -g openclaw

# Verify
openclaw --version
openclaw health check

Q: How do I migrate from OpenClaw to ZeroClaw? ZeroClaw supports direct config import:

zeroclaw migrate --from ~/.openclaw
# Memory and config files are compatible; most skills work via MCP bridge

🤝 Community & Support

Official Channels

Channel Description
GitHub Issues Bug reports, feature feedback, CVE disclosures
Discord Real-time help; #skills, #deployment, #security channels
X / Twitter News, tips, community highlights

Key People to Follow

Handle Why
@steipete Creator; shares architectural thinking
@voltagent_dev Maintains awesome-openclaw-skills
@KSimback Best ongoing alternatives ecosystem analysis
@adversa_ai Security researchers behind SecureClaw

Newsletters

Newsletter Focus
OpenClaw Newsletter Daily updates: releases, community highlights, security alerts and ecosystem news, curated by community maintainers on Buttondown

🗺️ Related Awesome Lists

List Scope
VoltAgent/awesome-openclaw-skills 5,400+ curated skills — the specialized skills index
hesamsheikh/awesome-openclaw-usecases Real-world use cases by domain
vincentkoc/awesome-openclaw General community list
awesome-mcp-servers Full MCP ecosystem
awesome-ai-agents Broader AI agent landscape
awesome-llm-apps LLM application patterns
awesome-selfhosted Self-hosting ecosystem

🤝 Contributing

Contributions are welcome! This list aims to be opinionated and high-quality, not exhaustive.

Before submitting a PR:

  1. Have you actually used or read this resource? No link-farming.
  2. Add a description — every entry needs context: what, who, why.
  3. Check for duplicates first.
  4. Match the existing table/list format.
  5. Security resources get priority review.

See CONTRIBUTING.md for the full guide and CODE_OF_CONDUCT.md for community standards.

To report a malicious skill or security issue: Open an Issue with the security label. Do not submit CVEs as PRs.

Found this useful? Give it a ⭐ and share it with your team. Something missing or wrong? Open an issue or submit a PR.

Last updated: March 2026 · Maintained with ❤️ by the community

🌐 Language / 语言切换：English | 中文

About

A carefully curated list of awesome OpenClaw resources — not everything, just the best. Skills · Plugins · MCP · Tools · Deployments · Security · Research · Alternatives。精心策划的 OpenClaw 优质资源大合集 —— 不求大而全，只求真有用。

github.com/EthanYolo01/Awesome-OpenClaw

Topics

agent awesome awesome-list awesome-lists openclaw openclaw-agent

Resources

Readme

Code of conduct

Code of conduct

Contributing

Contributing

Security policy

Security policy
Activity

Stars

153 stars

Watchers

0 watching

Forks

18 forks
Report repository

Releases

No releases published

Packages

 
 
 

Contributors