openclaw-security-monitor

KO | EN

gitlite — search

#python #hacktoberfest #ai #typescript #android #rust #linux #nodejs #windows #cli #kubernetes #awesome-list #security #openai

openclaw-security-monitor

(★ 48)

Proactive security monitoring for OpenClaw deployments. Detects ClawHavoc, AMOS stealer, CVE-2026-25253, memory poisoning, and supply chain attacks.

openclaw-security-monitor 최신버젼 다운로드

최종 버전 다운로드 (.zip)

// repository documentation

Was this content helpful?

★ 0 (0 ratings)

# OpenClaw Security Monitor

Proactive security monitoring, threat scanning, and real-time visibility for [OpenClaw](https://github.com/openclawai/openclaw) deployments. Detects threats from the **ClawHavoc** campaign (824+ malicious skills), **AMOS stealer**, **Vidar infostealer**, **GhostSocks** proxy malware, **ClawJacked** WebSocket brute-force, workspace plugin auto-loading attacks, shared-auth scope escalation, approval replay/integrity bypasses, supply chain attacks, memory poisoning, log poisoning, browser relay hijacking, TAR traversal, SSRF, SHA-1 cache poisoning, MCP tool poisoning, SANDWORM worm propagation, **170 advisories**, and **500+ CVEs**.

## Why This Exists

In late January 2026, security researchers found that **12% of all ClawHub skills were malicious** — 341 out of 2,857 skills across multiple campaigns. By mid-February, this expanded to **824+ malicious skills** with **1,184 malicious packages** across 12 publisher accounts (Antiy CERT). The Snyk ToxicSkills study found **36% of all ClawHub skills contain security flaws** (3,984 scanned).

The primary campaign, ClawHavoc, delivered the Atomic Stealer (AMOS) macOS infostealer targeting crypto wallets, SSH credentials, and browser passwords. In February, Hudson Rock discovered **Vidar infostealer variants specifically targeting OpenClaw agent identities** — stealing openclaw.json, device.json, soul.md, and memory.md files.

Meanwhile, CVE-2026-25253 demonstrated that a single malicious link could achieve full remote code execution on any OpenClaw instance through WebSocket hijacking — even those bound to localhost. The **ClawJacked** attack (Feb 26, Oasis Security) showed that malicious websites can brute-force localhost WebSocket passwords with no rate limiting. **CVE-2026-28363** (CVSS 9.9) revealed a critical safeBins bypass via GNU long-option abbreviations. In total, **170 advisories and 500+ CVEs** have been tracked across project and third-party disclosures including SSRF, exec bypass, ACP auto-approval bypass, webhook forgery, log poisoning, and more. The March 19-21 batch (CVE-2026-32013, CVE-2026-32014, CVE-2026-32025, CVE-2026-32042, CVE-2026-32048, CVE-2026-32051, CVE-2026-32055, CVE-2026-32056, CVE-2026-32064) added symlink traversal, sandbox escape, shell environment RCE, unauthenticated VNC observer access, and device identity/metadata spoofing. The April 16 wave added **Matrix room-control sender-auth bypass** (GHSA-2gvc-4f3c-2855), **webchat media local-root bypass** (GHSA-mr34-9552-qr95), **gateway SecretRef stale bearer auth** (GHSA-xmxx-7p24-h892), and **config.get redaction bypass** (GHSA-8372-7vhw-cm6q). The April 21-25 rollups added **setup-api.js cwd execution** (GHSA-r39h-4c2p-3jxp), **webhook SecretRef route-secret replay** (GHSA-q8ff-7ffm-m3r9), **gateway config mutation guard bypass** (GHSA-cwj3-vqpp-pmxr), dotenv connector/runtime overrides, MCP owner-context/tool-policy issues, OpenShell FS bridge escapes, and additional SSRF/media hardening. The June 2026 rollup added **MCP Streamable HTTP redirect header leakage** (CVE-2026-53840), **workspace dotenv/env command influence** (CVE-2026-53842, CVE-2026-53846, CVE-2026-53858, CVE-2026-53864), **exec allowlist bypasses** (CVE-2026-53853, CVE-2026-53855, CVE-2026-53861, CVE-2026-53866), **service PATH hijacking** (CVE-2026-53865), and **revoked device-token/scope abuse** (CVE-2026-53843, CVE-2026-53847, CVE-2026-53852). Unit 42 also documented five evasive ClawHub skills from February-May 2026, including TradingView paste-site lures, an oversized padded README dropper (`omnicogg`), runtime affiliate injection (`money-radar`), and agentic front-running (`letssendit`).

**135,000+ instances** are exposed across 82 countries, with **12,812 exploitable via RCE**. Major security firms including CrowdStrike, Bitdefender, Palo Alto Networks, Cisco, and Kaspersky have issued advisories. Meta has banned OpenClaw from corporate devices.

This project provides defense-in-depth monitoring for self-hosted OpenClaw installations. **Minimum safe version: v2026.5.26**.

## Features

- **41-point security scan** covering C2 infrastructure, stealers, reverse shells, credential exfiltration, memory poisoning, SKILL.md injection, WebSocket hijacking, ClawJacked brute-force, SSRF, safeBins bypass, ACP auto-approval, PATH hijacking, env override injection, deep link truncation, log poisoning, SHA-1 cache poisoning, Google Chat webhook bypass, CSWSH, MCP tool poisoning, SANDWORM worm detection, rules file backdoor, setup-api.js plugin hijacks, workspace plugin auto-loading, shared-auth scope abuse, exec approval replay, file-padding evasion, malicious skill-name patterns, DM/tool/sandbox policies, persistence mechanisms, plugin auditing, Docker security, and more
- **IOC database** with known C2 IPs, malicious domains, file hashes, publisher blacklists, and skill name patterns
- **Auto-updating IOC feeds** that pull latest threat intelligence from upstream
- **Bundle-plugin manifest** for `clawhub package publish --family bundle-plugin`
- **Web dashboard** (dark-themed, zero dependencies) with real-time status, process trees, network monitoring, and scan history
- **Daily automated scans** with Telegram alerting
- **Process ancestry tracking** via [witr](https://github.com/pranshuparmar/witr) integration

## Quick Start

```bash
# Clone into OpenClaw skills directory
git clone https://github.com/adibirzu/openclaw-security-monitor.git \
  ~/.openclaw/workspace/skills/openclaw-security-monitor

cd ~/.openclaw/workspace/skills/openclaw-security-monitor

# Make scripts executable
chmod +x scripts/*.sh scripts/remediate/*.sh

# Run a 41-point scan (read-only, no system changes)
./scripts/scan.sh

# Preview what remediation would do (dry-run, no changes)
./scripts/remediate.sh --dry-run

# Start the web dashboard (read-only, localhost:18800)
node dashboard/server.js

# Scan all installed ClawHub skills
./scripts/clawhub-scan.sh

# Update IOC database (interactive, asks for confirmation)
./scripts/update-ioc.sh
```

## Bundle Plugin

This project is also publishable as a ClawHub bundle plugin.

```bash
# Install the bundle plugin variant
openclaw plugins install clawhub:openclaw-security-monitor-bundle
```

**Optional persistence** (manual, not auto-installed):
```bash
# Install daily cron (06:00 UTC) — requires explicit user action
crontab -l | { cat; echo "0 6 * * * $(pwd)/scripts/daily-scan-cron.sh"; } | crontab -
```

## Architecture

```
openclaw-security-monitor/
  .codex-plugin/
    plugin.json          # Bundle-plugin manifest for ClawHub/OpenClaw plugin publishing
  scripts/
    scan.sh              # 41-point threat scanner (v5.5.0)
    remediate.sh         # Orchestrator: scan + per-check remediation
    remediate/
      _common.sh         # Shared helpers (log, confirm, fix_perms)
      check-01-c2-ips.sh ... check-41-device-identity-spoofing.sh  # 41 consolidated per-check scripts
      check-42-*.sh ... check-62-*.sh  # retained legacy advisory-specific helpers
    clawhub-scan.sh      # Scan all installed ClawHub skills against IOC database
    dashboard.sh         # CLI security dashboard with witr
    network-check.sh     # Network activity monitor
    daily-scan-cron.sh   # Cron wrapper + Telegram alerts
    telegram-setup.sh    # Telegram notification setup
    update-ioc.sh        # IOC database updater
  ioc/
    c2-ips.txt           # Known C2 IP addresses
    malicious-domains.txt # Payload/exfil domains
    file-hashes.txt      # Known malicious file hashes
    malicious-publishers.txt  # Blacklisted ClawHub accounts
    malicious-skill-patterns.txt  # Malicious skill naming patterns
  dashboard/
    server.js            # Node.js HTTP server (zero npm deps)
    index.html           # Single-file dark-themed SPA
  docs/
    threat-model.md      # Threat model and attack vectors
```

## Scan Checks (41)

| # | Check | Severity | Detects |
|---|-------|----------|---------|
| 1 | C2 Infrastructure | CRITICAL | Known C2 IPs in skill code |
| 2 | Malware Signatures & Obfuscation | CRITICAL | AMOS stealer, base64 obfuscation, binary downloads, file-padding evasion |
| 3 | Reverse Shells | CRITICAL | bash/python/perl/ruby/php/lua reverse shells |
| 4 | Credential Exfiltration | CRITICAL | webhook.site, pipedream, ngrok, burpcollaborator |
| 5 | Crypto Wallet Targeting | WARNING | Seed phrases, private keys, exchange API keys |
| 6 | Curl-Pipe Attacks | WARNING | curl\|sh, wget\|bash, remote script execution |
| 7 | File & Credential Permission Audit | WARNING | Config files, credentials dir, session perms |
| 8 | Skill Integrity | WARNING | SKILL.md hash changes since last scan |
| 9 | AI Prompt Injection & Instruction Manipulation | CRITICAL | SKILL.md injection, memory poisoning, MCP tool poisoning, rules file backdoor, prompt-channel trust fixes |
| 10 | Gateway Config | CRITICAL | Auth disabled, LAN exposure, version check |
| 11 | WebSocket Security | CRITICAL | CVE-2026-25253, ClawJacked, device identity skip, CSWSH |
| 12 | Malicious Publishers | CRITICAL | Skills from known-bad ClawHub accounts and installed skill-name patterns |
| 13 | Credential Leakage & Plaintext Secrets | WARNING | Env leakage, hardcoded API keys, plaintext credentials |
| 14 | DM, Tool & Sandbox Policies | CRITICAL | Open DM, wildcard tools, disabled sandbox, wildcard owner-command bypasses |
| 15 | mDNS/Bonjour Exposure | WARNING | mDNS broadcasting in full mode |
| 16 | Persistence Mechanisms | WARNING | Unauthorized LaunchAgents, crontabs, systemd |
| 17 | Log Security & Poisoning | WARNING | Redaction disabled, ANSI injection, header injection |
| 18 | Plugin/Extension Audit | CRITICAL | Extensions with exec patterns, malicious domains, setup-api.js cwd hijacks |
| 19 | Docker Security | CRITICAL | Root containers, socket mount, privileged mode |
| 20 | Authentication & Route Security | CRITICAL | Proxy bypass, CDP auth, browser bridge, /agent/act, SecretRef replay, config mutation guard, June auth/scope CVEs |
| 21 | Exec Guardrails & Approval Security | CRITICAL | safeBins bypass, shell expansion, field injection, heredoc/env/applet gaps, replay, June exec allowlist CVEs |
| 22 | Node.js CVE Check | WARNING | CVE-2026-21636 permission model bypass |
| 23 | VS Code Trojans | CRITICAL | Fake ClawdBot/OpenClaw VS Code extensions |
| 24 | Internet Exposure | WARNING | Non-loopback gateway binding |
| 25 | MCP Server Security | CRITICAL | Unrestricted MCP servers, prompt injection, env poisoning, owner-context/tool-policy bypass, CVE-2026-53840 redirects |
| 26 | PATH Hijacking & Command Resolution | CRITICAL | GHSA-jqpq, CVE-2026-29610, CVE-2026-53865 command hijacking |
| 27 | SSRF Protection | WARNING | CVE-2026-26322, CVE-2026-27488, QQBot/Zalo/browser media SSRF, CVE-2026-53859 hostname checks |
| 28 | Path Traversal & File Handling | CRITICAL | Deep link truncation, browser control, TAR traversal, OpenShell FS bridge escapes |
| 29 | DoS Protection | WARNING | CVE-2026-28478, CVE-2026-29609 memory exhaustion |
| 30 | ACP Auto-Approval | WARNING | GHSA-7jx5 untrusted metadata bypass |
| 31 | Env Override Injection | WARNING | GHSA-82g8 skill env overrides, workspace dotenv connector/runtime overrides, June env CVEs |
| 32 | Privilege Escalation & Scope Abuse | CRITICAL | Pairing creds, operator escalation, shared-auth, paired-device, node-token, active-memory and ACP scope abuse |
| 33 | SHA-1 Cache Poisoning | CRITICAL | CVE-2026-28479 SHA-1 collision cache attack |
| 34 | Google Chat Webhook Bypass | CRITICAL | CVE-2026-28469 cross-account webhook injection |
| 35 | SANDWORM Worm Detection | CRITICAL | Autonomous MCP worm propagation |
| 36 | Workspace Plugin Auto-Discovery | CRITICAL | GHSA-99qw malicious .openclaw/extensions loading |
| 37 | Symlink Traversal | CRITICAL | CVE-2026-32013, CVE-2026-32055 symlink escape |
| 38 | Sandbox Escape & Session Inheritance | CRITICAL | CVE-2026-32048, CVE-2026-32051 session bypass |
| 39 | Shell Environment RCE | CRITICAL | CVE-2026-32056, CVE-2026-27566 env injection |
| 40 | VNC & Observer Authentication | CRITICAL | CVE-2026-32064 unauthenticated VNC observer |
| 41 | Device Identity & Metadata Spoofing | CRITICAL | CVE-2026-32014, CVE-2026-32042, CVE-2026-32025 |

## Remediation Guide

Step-by-step hardening for each security check, covering both macOS and Linux.

### Check 1: C2 Infrastructure — Known C2 IPs in skill code

**Severity:** CRITICAL

**What it means:** A skill contains IP addresses associated with known command-and-control servers (e.g., `91.92.242.30`). This strongly indicates the skill is malicious.

**Remediation:**
```bash
# Identify the affected skill
grep -rlE "91\.92\.242|95\.92\.242|54\.91\.154\.110" ~/.openclaw/workspace/skills/

# Remove the malicious skill
openclaw skill remove <skill-name>

# Or manually delete it
rm -rf ~/.openclaw/workspace/skills/<skill-name>

# Verify removal
grep -rlE "91\.92\.242" ~/.openclaw/workspace/skills/ # should return empty
```

### Check 2: AMOS Stealer / AuthTool Markers

**Severity:** CRITICAL

**What it means:** A skill contains patterns associated with the Atomic Stealer (AMOS) macOS infostealer, NovaStealer, or the "AuthTool" social engineering binary.

**Remediation:**
```bash
# Remove the malicious skill immediately
openclaw skill remove <skill-name>

# Check if AMOS was executed (macOS)
# Look for suspicious LaunchAgents
ls ~/Library/LaunchAgents/ | grep -ivE "com\.apple|com\.openclaw\.security"

# Check for unexpected login items (macOS)
osascript -e 'tell application "System Events" to get the name of every login item'

# Check browser extensions were not tampered with
ls ~/Library/Application\ Support/Google/Chrome/Default/Extensions/

# If AMOS was executed: rotate ALL credentials (SSH, AWS, crypto wallets, browser passwords)
```

### Check 3: Reverse Shells & Backdoors

**Severity:** CRITICAL

**What it means:** A skill contains reverse shell patterns (`nc -e`, `/dev/tcp/`, `socat exec`, etc.) or Gatekeeper bypass commands (`xattr -cr`).

**Remediation:**
```bash
# Remove the skill
openclaw skill remove <skill-name>

# Kill any suspicious connections
kill -9 <PID>

# Re-enable Gatekeeper if bypassed (macOS)
sudo spctl --master-enable
sudo defaults write com.apple.LaunchServices LSQuarantine -bool true

# Check for quarantine removals (macOS)
xattr -l /Applications/*.app | grep quarantine
```

### Check 4: Credential Exfiltration Endpoints

**Severity:** CRITICAL

**What it means:** A skill sends data to known exfiltration services (webhook.site, pipedream.net, ngrok.io, etc.).

**Remediation:**
```bash
# Remove the skill
openclaw skill remove <skill-name>

# Block exfiltration domains at the host level
# macOS/Linux:
sudo sh -c 'echo "127.0.0.1 webhook.site" >> /etc/hosts'
sudo sh -c 'echo "127.0.0.1 pipedream.net" >> /etc/hosts'
sudo sh -c 'echo "127.0.0.1 hookbin.com" >> /etc/hosts'
sudo sh -c 'echo "127.0.0.1 requestbin.com" >> /etc/hosts'

# Flush DNS cache (macOS)
sudo dscacheutil -flushcache && sudo killall -HUP mDNSResponder
# Flush DNS cache (Linux)
sudo systemd-resolve --flush-caches
```

### Check 5: Crypto Wallet Targeting

**Severity:** WARNING

**What it means:** A skill references crypto wallet private keys, seed phrases, or exchange API keys.

**Remediation:**
```bash
# Remove the skill
openclaw skill remove <skill-name>

# Move crypto wallets to hardware wallets (Ledger, Trezor)
# Rotate exchange API keys immediately

# Restrict filesystem access (macOS)
chmod 700 ~/Library/Application\ Support/Phantom
chmod 700 ~/Library/Application\ Support/MetaMask

# Check if wallet files were accessed recently
# macOS:
mdls -name kMDItemLastUsedDate ~/Library/Application\ Support/Phantom/*
# Linux:
stat -c "%x" ~/.config/phantom/*
```

### Check 6: Curl-Pipe Attacks

**Severity:** WARNING

**What it means:** A skill uses `curl | sh` or `wget | bash` patterns to download and execute remote scripts.

**Remediation:**
```bash
# Remove the skill
openclaw skill remove <skill-name>

# Review and restrict agent tool access
openclaw config set tools.deny '["exec","process"]'

# Block curl-pipe at the shell level (add to ~/.bashrc or ~/.zshrc)
# This creates an alias that warns on pipe-to-shell patterns:
alias curl='() { if [[ "$*" == *"|"*"sh"* ]] || [[ "$*" == *"|"*"bash"* ]]; then echo "BLOCKED: curl-pipe detected"; return 1; fi; command curl "$@"; }'
```

### Check 7: File Permission Audit

**Severity:** WARNING

**What it means:** Sensitive configuration files (openclaw.json, auth-profiles.json) are readable by other users.

**Remediation:**
```bash
# Fix permissions (macOS/Linux)
chmod 600 ~/.openclaw/openclaw.json
chmod 600 ~/.openclaw/agents/main/agent/auth-profiles.json
chmod 600 ~/.openclaw/exec-approvals.json
chmod 700 ~/.openclaw

# Set default umask in shell profile (add to ~/.bashrc or ~/.zshrc)
echo 'umask 077' >> ~/.zshrc
```

### Check 8: Skill Integrity Hashes

**Severity:** WARNING

**What it means:** SKILL.md files have been modified since the last scan, which could indicate tampering or supply chain compromise.

**Remediation:**
```bash
# Review what changed
diff ~/.openclaw/logs/skill-hashes.sha256.prev ~/.openclaw/logs/skill-hashes.sha256

# For each changed skill, inspect the diff
cd ~/.openclaw/workspace/skills/<skill-name>
git diff HEAD~1 SKILL.md   # If under version control

# If change is unexpected, reinstall from ClawHub
openclaw skill remove <skill-name>
openclaw skill install <skill-name>

# Enable skill pinning (if supported)
openclaw config set skills.autoUpdate false
```

### Check 9: SKILL.md Shell Injection

**Severity:** WARNING

**What it means:** A SKILL.md contains suspicious install instructions that try to get the user or agent to run shell commands (Snyk CVE-2026-22708).

# Remove the skill if instructions are clearly malicious
openclaw skill remove <skill-name>

# Never run commands from SKILL.md Prerequisites without reviewing them first
# Configure agent to not auto-execute installation commands
openclaw config set tools.deny '["exec"]'
```

### Check 10: Memory Poisoning

**Severity:** CRITICAL

**What it means:** SOUL.md, MEMORY.md, or IDENTITY.md contain instruction-override patterns (for example, role-reset language that attempts to redefine agent behavior).

**Remediation:**
```bash
# Review poisoned files
cat ~/.openclaw/workspace/SOUL.md
cat ~/.openclaw/workspace/MEMORY.md
cat ~/.openclaw/workspace/IDENTITY.md

# Remove injected content manually (edit the file, remove injected lines)
# Or restore from backup:
git checkout HEAD -- ~/.openclaw/workspace/SOUL.md

# Protect memory files from modification
chmod 444 ~/.openclaw/workspace/SOUL.md
chmod 444 ~/.openclaw/workspace/MEMORY.md
chmod 444 ~/.openclaw/workspace/IDENTITY.md

# Investigate which skill wrote to memory
grep -rl "SOUL\.md\|MEMORY\.md" ~/.openclaw/workspace/skills/
```

### Check 11: Base64 Obfuscation

**Severity:** WARNING

**What it means:** A skill uses base64 encoding/decoding, which is a common technique to hide malicious payloads (as seen in the ClawHavoc campaign via glot.io).

**Remediation:**
```bash
# Review the base64 content
grep -rn "base64" ~/.openclaw/workspace/skills/<skill-name>/

# Decode and inspect the payload
echo "<base64-string>" | base64 -d

# Remove if malicious
openclaw skill remove <skill-name>

# Monitor for base64 activity in agent logs
grep -i "base64" ~/.openclaw/logs/*.log
```

### Check 12: External Binary Downloads

**Severity:** WARNING

**What it means:** A skill references downloadable binaries (.exe, .dmg, .pkg, .zip with password) or known malicious download URLs.

**Remediation:**
```bash
# Remove the skill
openclaw skill remove <skill-name>

# Check Downloads folder for suspicious binaries (macOS)
ls -la ~/Downloads/*.{exe,dmg,pkg,zip,msi} 2>/dev/null

# Verify Gatekeeper is active (macOS)
spctl --status  # should say "assessments enabled"

# Check if any unsigned apps were installed (macOS)
sudo find /Applications -name "*.app" -exec codesign -v {} \; 2>&1 | grep "invalid"

# Linux: check /tmp and /var/tmp for suspicious downloads
find /tmp /var/tmp -name "*.sh" -o -name "*.elf" -o -name "*.bin" -mtime -7 2>/dev/null
```

### Check 13: Gateway Security Configuration

**Severity:** CRITICAL

**What it means:** The OpenClaw gateway is bound to LAN (accessible from network) or has authentication disabled.

**Remediation:**
```bash
# Bind gateway to localhost only
openclaw config set gateway.bind localhost

# Enable authentication
openclaw config set gateway.auth.mode token
# Or use OIDC (enterprise)
openclaw config set gateway.auth.mode oidc

# Verify configuration
openclaw config get gateway.bind
openclaw config get gateway.auth.mode

# Check what's actually listening
lsof -i :18789 -nP   # Should show 127.0.0.1, not 0.0.0.0

# Update to latest version to patch CVE-2026-25253
openclaw update
```

### Check 14: WebSocket Security (CVE-2026-25253)

**Severity:** CRITICAL

**What it means:** The gateway WebSocket accepts connections from arbitrary origins, enabling 1-click RCE via a malicious webpage.

**Remediation:**
```bash
# Update OpenClaw to latest version (patch included in 2026.2.3+)
openclaw update

# Verify the fix
curl -s -o /dev/null -w "%{http_code}" \
  -H "Connection: Upgrade" -H "Upgrade: websocket" \
  -H "Sec-WebSocket-Key: dGhlIHNhbXBsZSBub25jZQ==" \
  -H "Sec-WebSocket-Version: 13" \
  -H "Origin: http://evil.attacker.com" \
  http://127.0.0.1:18789/
# Should return 403 or 401, NOT 101

# If update is not available, restrict gateway to localhost
openclaw config set gateway.bind localhost

# Add firewall rule (macOS)
sudo /usr/libexec/ApplicationFirewall/socketfilterfw --add /usr/local/bin/node
# Linux (iptables)
sudo iptables -A INPUT -p tcp --dport 18789 -s 127.0.0.1 -j ACCEPT
sudo iptables -A INPUT -p tcp --dport 18789 -j DROP
```

### Check 15: Malicious Publisher Detection

**Severity:** CRITICAL

**What it means:** An installed skill references a known malicious ClawHub publisher (hightower6eu, zaycv, Ddoy233, etc.).

**Remediation:**
```bash
# Remove all skills from the malicious publisher
openclaw skill remove <skill-name>

# Check for other skills from the same publisher
grep -rl "<publisher-name>" ~/.openclaw/workspace/skills/

# Review recently installed skills
ls -lt ~/.openclaw/workspace/skills/ | head -20

# Only install skills from verified publishers
# Check skill reputation before installing:
# https://www.koi.ai (Clawdex reputation checker)
```

### Check 16: Sensitive Environment Leakage

**Severity:** WARNING / CRITICAL (if hardcoded API keys found)

**What it means:** A skill reads sensitive files (.env, .ssh, .aws/credentials) or contains hardcoded API keys / Moltbook tokens.

**Remediation:**
```bash
# Remove the skill
openclaw skill remove <skill-name>

# Rotate exposed credentials immediately
# OpenAI:
# Go to https://platform.openai.com/api-keys and regenerate
# Anthropic:
# Go to https://console.anthropic.com/settings/keys and regenerate

# Move secrets to a secrets manager instead of .env files
# macOS: Use Keychain Access
security add-generic-password -a "$USER" -s "OPENAI_API_KEY" -w "<new-key>"

# Linux: Use pass or secret-tool
secret-tool store --label="OPENAI_API_KEY" service openai key api

# Restrict skill filesystem access
openclaw config set sandbox.mode all
```

### Check 17: DM Policy Audit

**Severity:** WARNING

**What it means:** A messaging channel has `dmPolicy=open`, allowing anyone to message your agent — enabling social engineering and prompt injection from untrusted sources.

**Remediation:**
```bash
# Set DM policy to restricted for each channel
openclaw config set channels.whatsapp.dmPolicy restricted
openclaw config set channels.telegram.dmPolicy restricted
openclaw config set channels.discord.dmPolicy restricted

# Set explicit allowFrom list
openclaw config set channels.telegram.allowFrom '["your-user-id"]'

# Disable unused channels entirely
openclaw config set channels.signal.enabled false
```

### Check 18: Tool Policy / Elevated Tools Audit

**Severity:** CRITICAL

**What it means:** Elevated tools are enabled with wildcard access, or the tool deny list is empty — allowing the agent to execute arbitrary commands.

**Remediation:**
```bash
# Restrict elevated tools to specific principals
openclaw config set tools.elevated.allowFrom '["your-user-id"]'

# Add dangerous tools to the deny list
openclaw config set tools.deny '["exec","process","browser","filesystem-write"]'

# Require approval for elevated actions
openclaw config set tools.elevated.requireApproval true

# Review current tool permissions
openclaw config get tools
```

### Check 19: Sandbox Configuration

**Severity:** WARNING

**What it means:** Sandboxing is disabled, allowing skills to access the full filesystem and network without restriction.

**Remediation:**
```bash
# Enable sandboxing
openclaw config set sandbox.mode all

# Restrict workspace access to read-only where possible
openclaw config set sandbox.workspaceAccess ro

# For Docker deployments (recommended for production):
# Use gVisor runtime for stronger isolation
docker run --runtime=runsc --read-only \
  -v ~/.openclaw:/home/openclaw/.openclaw:ro \
  openclaw/gateway
```

### Check 20: mDNS/Bonjour Exposure

**Severity:** WARNING

**What it means:** mDNS is broadcasting in "full" mode, advertising the gateway's presence, paths, and SSH port to the local network.

**Remediation:**
```bash
# Disable mDNS broadcasting
openclaw config set discovery.mdns.mode off

# Or set to minimal (name only, no paths)
openclaw config set discovery.mdns.mode minimal

# Verify mDNS is not broadcasting (macOS)
dns-sd -B _openclaw._tcp local
# Should return no results

# Disable Bonjour for the gateway specifically (macOS)
sudo defaults write /Library/Preferences/com.apple.mDNSResponder.plist NoMulticastAdvertisements -bool YES
```

### Check 21: Session & Credential Permissions

**Severity:** WARNING

**What it means:** Credential directories, session files, or the OpenClaw home directory have overly permissive permissions.

**Remediation:**
```bash
# Fix all permissions at once
chmod 700 ~/.openclaw
chmod 700 ~/.openclaw/credentials 2>/dev/null
chmod 700 ~/.openclaw/agents/*/sessions 2>/dev/null
find ~/.openclaw/credentials -type f -name "*.json" -exec chmod 600 {} \; 2>/dev/null
find ~/.openclaw -name "auth-profiles.json" -exec chmod 600 {} \;

# Verify
ls -la ~/.openclaw/
ls -la ~/.openclaw/credentials/
```

### Check 22: Persistence Mechanisms

**Severity:** WARNING

**What it means:** There are LaunchAgents, cron entries, or systemd services referencing OpenClaw that are not the known security monitor.

**Remediation:**
```bash
# macOS: Review LaunchAgents
ls -la ~/Library/LaunchAgents/ | grep -iE "openclaw|clawdbot|moltbot"
# Remove unauthorized agents:
launchctl unload ~/Library/LaunchAgents/<suspicious-plist>
rm ~/Library/LaunchAgents/<suspicious-plist>

# Check system-level LaunchDaemons (requires root)
sudo ls /Library/LaunchDaemons/ | grep -iE "openclaw|clawdbot"

# Linux: Review systemd services
systemctl --user list-units --type=service | grep -iE "openclaw|clawdbot"
# Disable unauthorized services:
systemctl --user disable --now <service-name>

# Review crontab
crontab -l | grep -iE "openclaw|clawdbot|moltbot"
# Edit and remove suspicious entries:
crontab -e
```

### Check 23: Plugin/Extension Security

**Severity:** CRITICAL

**What it means:** An installed extension contains code execution patterns (`eval()`, `exec()`, `child_process`) or references known malicious domains.

**Remediation:**
```bash
# List installed extensions
ls ~/.openclaw/extensions/

# Review suspicious extension code
grep -rn "eval\|exec\|child_process\|fetch(" ~/.openclaw/extensions/<ext-name>/

# Remove the extension
rm -rf ~/.openclaw/extensions/<ext-name>

# Disable extension loading if not needed
openclaw config set extensions.enabled false

# Only install extensions from verified sources
```

### Check 24: Log Redaction Audit

**Severity:** WARNING

**What it means:** Log redaction is disabled (sensitive data like API keys and passwords may appear in plaintext in logs), or log directories are world-readable.

**Remediation:**
```bash
# Enable log redaction
openclaw config set logging.redactSensitive true

# Fix log directory permissions
chmod 700 ~/.openclaw/logs
chmod 700 /tmp/openclaw 2>/dev/null

# Verify logs don't contain sensitive data
grep -iE "sk-[a-zA-Z0-9]{10}|password|token.*=" ~/.openclaw/logs/*.log

# Set up log rotation (macOS - newsyslog)
sudo sh -c 'echo "$HOME/.openclaw/logs/*.log 640 7 1000 * J" >> /etc/newsyslog.conf'
# Linux - logrotate
cat > /etc/logrotate.d/openclaw << 'LOGROTATE'
/home/*/.openclaw/logs/*.log {
    weekly
    rotate 4
    compress
    missingok
    notifempty
    create 0600 root root
}
LOGROTATE
```

### Check 25: Reverse Proxy Localhost Trust Bypass

**Severity:** CRITICAL

**What it means:** The gateway is bound to LAN without `trustedProxies` configured, or `dangerouslyDisableDeviceAuth` is enabled. External attackers can bypass authentication by appearing as localhost through a reverse proxy (28% of exposed instances had this flaw per Penligent).

**Remediation:**
```bash
# Configure trusted proxies
openclaw config set gateway.trustedProxies '["192.168.1.1"]'

# Never disable device auth
openclaw config set gateway.dangerouslyDisableDeviceAuth false

# If using Nginx, ensure X-Forwarded-For is properly set
# In nginx.conf:
# proxy_set_header X-Real-IP $remote_addr;
# proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;

# Verify
openclaw config get gateway.trustedProxies
```

### Check 26: Exec-Approvals Configuration

**Severity:** CRITICAL

**What it means:** The `exec-approvals.json` allows remote command execution without confirmation. The CVE-2026-25253 exploit chain used `exec.approvals.set` to disable confirmation prompts.

**Remediation:**
```bash
# Set exec approvals to deny for non-owner nodes
# Edit ~/.openclaw/exec-approvals.json manually:
# Set "security": "deny" for all non-owner nodes
# Set "ask": "always"

# Fix permissions
chmod 600 ~/.openclaw/exec-approvals.json

# Verify
cat ~/.openclaw/exec-approvals.json | python3 -m json.tool
```

### Check 27: Docker Container Security

**Severity:** CRITICAL

**What it means:** OpenClaw Docker containers are running as root, with Docker socket mounted, or in privileged mode. CVE-2026-24763 was a command injection in the Docker sandbox.

**Remediation:**
```bash
# Run OpenClaw container with hardened settings
docker run -d \
  --name openclaw \
  --user 1000:1000 \
  --cap-drop ALL \
  --no-new-privileges \
  --read-only \
  --tmpfs /tmp:noexec,nosuid,size=100m \
  --memory 512m \
  --cpus 1.0 \
  --pids-limit 100 \
  -v ~/.openclaw:/home/openclaw/.openclaw:rw \
  openclaw/gateway

# NEVER mount Docker socket
# BAD: -v /var/run/docker.sock:/var/run/docker.sock

# Use gVisor for stronger isolation
docker run --runtime=runsc ...
```

### Check 28: Node.js Version / CVE-2026-21636

**Severity:** WARNING

**What it means:** Node.js version is below 22.12.0 and vulnerable to CVE-2026-21636 (permission model bypass via Unix Domain Sockets), which allows sandbox escape.

**Remediation:**
```bash
# macOS
brew install node@22
# Or update existing
brew upgrade node

# Linux (NodeSource)
curl -fsSL https://deb.nodesource.com/setup_22.x | sudo bash -
sudo apt-get install -y nodejs

# Verify
node --version  # Should be >= v22.12.0
```

### Check 29: Plaintext Credential Detection

**Severity:** WARNING

**What it means:** OpenClaw config or credential files contain plaintext API keys (OpenAI `sk-`, AWS `AKIA`, GitHub `ghp_`, Slack `xoxb-`). The `~/.openclaw` directory is a standard target for infostealers.

**Remediation:**
```bash
# macOS: Move secrets to Keychain
security add-generic-password -a "$USER" -s "OPENAI_API_KEY" -w "<key>"

# Linux: Use pass or secret-tool
pass insert openai/api-key
# Or
secret-tool store --label="OPENAI_API_KEY" service openai key api

# Use environment variables instead of config files
export OPENAI_API_KEY=$(security find-generic-password -s "OPENAI_API_KEY" -w)

# Set up detect-secrets baseline
pip install detect-secrets
detect-secrets scan ~/.openclaw/ > .secrets.baseline
```

### Check 30: VS Code Extension Trojan Detection

**Severity:** CRITICAL

**What it means:** Fake "ClawdBot Agent" VS Code extensions have been discovered that install ScreenConnect RAT and remote access trojans. OpenClaw has NO official VS Code extension.

**Remediation:**
```bash
# Remove any clawdbot/openclaw VS Code extensions immediately
rm -rf ~/.vscode/extensions/*clawdbot*
rm -rf ~/.vscode/extensions/*openclaw*
rm -rf ~/.vscode/extensions/*moltbot*

# Check for ScreenConnect/ConnectWise artifacts
# macOS:
find /Applications -name "*ScreenConnect*" -o -name "*ConnectWise*"
# Linux:
find /opt -name "*screenconnect*" -o -name "*connectwise*"

# Scan for persistence left by the trojan
ls ~/Library/LaunchAgents/ | grep -ivE "com\.apple"
```

### Check 31: Internet Exposure Detection

**Severity:** WARNING

**What it means:** The gateway is listening on non-loopback interfaces, potentially exposing it to the internet. Over 21,000 OpenClaw instances were found on Shodan with zero authentication.

**Remediation:**
```bash
# Bind to localhost only
openclaw config set gateway.bind localhost

# Verify
lsof -i :18789 -nP | grep LISTEN
# Should show 127.0.0.1:18789, NOT *:18789 or 0.0.0.0:18789

# Add firewall rules (macOS pf)
echo "block in on ! lo0 proto tcp to any port 18789" | sudo pfctl -ef -

# Linux (iptables)
sudo iptables -A INPUT -p tcp --dport 18789 ! -s 127.0.0.1 -j DROP

# Change default port (reduce scan surface)
openclaw config set gateway.port 28789
```

### Check 32: MCP Server Security

**Severity:** CRITICAL

**What it means:** MCP servers are configured with unrestricted access (`enableAllProjectMcpServers=true`) or contain prompt injection patterns in tool descriptions — enabling tool poisoning and rug-pull attacks.

**Remediation:**
```bash
# Use explicit MCP server allowlist
openclaw config set mcp.enableAllProjectMcpServers false

# Review each MCP server's tool descriptions
cat ~/.openclaw/mcp.json | python3 -m json.tool

# Pin MCP server versions (prevent rug pulls)
# In mcp.json, use specific versions, not "latest"

# Run MCP servers in separate containers
# Use Cisco MCP Scanner for analysis:
# https://github.com/cisco-ai-defense/mcp-scanner

# Audit MCP server network access
# Check which external endpoints each server contacts
lsof -i -nP | grep -E "node|mcp" | grep ESTABLISHED
```

### Check 33: ClawJacked WebSocket Brute-Force (Oasis Security)

**Severity:** CRITICAL

**What it means:** OpenClaw's localhost WebSocket accepts connections without rate limiting, allowing brute-force attacks on the authentication password from any local process or malicious website.

**Remediation:**
```bash
# Update to v2026.2.25+ which adds rate limiting
openclaw update

# Verify version
openclaw --version  # Should be 2026.2.25 or later

# Enable strong authentication
openclaw config set gateway.auth.mode token
openclaw config set gateway.auth.tokenLength 64

# Monitor for brute-force attempts
grep "auth.*fail" ~/.openclaw/logs/*.log | tail -20
```

### Check 34: SSRF Protection (CVE-2026-26322, CVE-2026-27488)

**Severity:** WARNING

**What it means:** The gateway tool endpoint or cron webhook handler is vulnerable to server-side request forgery, allowing attackers to reach internal services.

**Remediation:**
```bash
# Update to v2026.2.10+ for SSRF patches
openclaw update

# Restrict outbound network access
openclaw config set network.allowedHosts '["api.anthropic.com","github.com"]'

# Block metadata endpoints (cloud environments)
sudo iptables -A OUTPUT -d 169.254.169.254 -j DROP  # AWS/GCP metadata
```

### Check 35: safeBins Bypass (CVE-2026-28363, CVSS 9.9)

**Severity:** CRITICAL

**What it means:** The exec safeBins allowlist can be bypassed using GNU long-option abbreviations (e.g., `sort --compress-prog=sh` executes arbitrary commands).

**Remediation:**
```bash
# Update to v2026.2.13+ which fixes the bypass
openclaw update

# Review exec-approvals for overly permissive entries
cat ~/.openclaw/exec-approvals.json | python3 -m json.tool

# Use full path binaries in allowlists instead of names
# Bad:  "sort" -> allows sort --compress-prog=sh
# Good: "/usr/bin/sort" with argument restrictions
```

### Check 36: ACP Auto-Approval Bypass (GHSA-7jx5)

**Severity:** WARNING

**What it means:** Untrusted SKILL.md metadata can bypass interactive tool approval prompts, allowing skills to execute tools without user consent.

**Remediation:**
```bash
# Update to v2026.2.12+ for the fix
openclaw update

# Disable auto-approval of ACP permissions
openclaw config set acp.autoApprove false

# Review which skills have ACP permissions
grep -rl "acp" ~/.openclaw/workspace/skills/*/SKILL.md
```

### Check 37: PATH Hijacking (GHSA-jqpq-mgvm-f9r6)

**Severity:** CRITICAL

**What it means:** Writable directories in the PATH before system directories allow attackers to plant malicious binaries that intercept OpenClaw exec calls.

**Remediation:**
```bash
# Check for writable dirs before system dirs
echo $PATH | tr ':' '\n' | while read d; do [ -w "$d" ] && echo "WRITABLE: $d"; done

# Remove writable dirs from PATH (add to ~/.zshrc)
export PATH="/usr/local/bin:/usr/bin:/bin:/usr/sbin:/sbin"

# Fix directory permissions
chmod 755 /usr/local/bin
chmod 755 ~/.local/bin
```

### Check 38: Skill Env Override Injection (GHSA-82g8-464f-2mv7)

**Severity:** WARNING

**What it means:** Skill environment variable overrides can redirect OpenClaw traffic to attacker-controlled hosts.

**Remediation:**
```bash
# Update to v2026.2.12+ for the fix
openclaw update

# Audit skill env overrides
grep -rn "env:" ~/.openclaw/workspace/skills/*/SKILL.md

# Remove skills with suspicious env overrides
openclaw skill remove <skill-name>
```

### Check 39: macOS Deep Link Truncation (CVE-2026-26320)

**Severity:** WARNING

**What it means:** macOS 240-character preview limit for deep links hides malicious payload portions from the user.

**Remediation:**
```bash
# Update to v2026.2.10+ for the fix
openclaw update

# Disable deep link handling if not needed
defaults write com.openclaw.gateway LSApplicationQueriesSchemes -array

# Always inspect full URLs before clicking deep links
```

### Check 40: Log Poisoning / WebSocket Header Injection

**Severity:** WARNING

**What it means:** WebSocket Origin/User-Agent headers can inject malicious content into agent logs, enabling indirect prompt injection when logs are reviewed by the agent.

**Remediation:**
```bash
# Update to latest version (includes log sanitization)
openclaw update

# Enable log redaction
openclaw config set logging.redactSensitive true

# Sanitize existing logs
sed -i.bak 's/\x1b\[[0-9;]*m//g' ~/.openclaw/logs/*.log

# Restrict log file permissions
chmod 600 ~/.openclaw/logs/*.log
```

### Check 41: Browser Relay CDP Unauthenticated Access (CVE-2026-28458)

**Severity:** CRITICAL

**What it means:** The /cdp WebSocket endpoint for browser control accepts connections without authentication, enabling session theft.

**Remediation:**
```bash
# Update to v2026.2.12+ for the fix
openclaw update

# Disable browser relay if not needed
openclaw config set browser.relay.enabled false

# Check for active CDP connections
lsof -i -nP | grep -E "cdp|devtools"
```

### Check 42: Browser Control Path Traversal (CVE-2026-28462)

**Severity:** CRITICAL

**What it means:** Path traversal in the browser control trace/download API allows arbitrary file writes outside the intended directory.

**Remediation:**
```bash
# Update to v2026.2.12+ for the fix
openclaw update

# Check for suspicious files written by browser control
find /tmp -name "*.trace" -newer /tmp -mtime -1 2>/dev/null
```

### Check 43: Shell Expansion Bypass (CVE-2026-28463)

**Severity:** CRITICAL

**What it means:** Exec-approvals validate commands before shell expansion but execute after expansion, allowing glob/variable injection to bypass allowlists.

**Remediation:**
```bash
# Update to v2026.2.14+ for the fix
openclaw update

# Use strict exec approval patterns (no wildcards)
# Review exec-approvals.json for entries containing * or $
grep -E '[\*\$]' ~/.openclaw/exec-approvals.json
```

### Check 44: Approval Field Injection (CVE-2026-28466)

**Severity:** CRITICAL

**What it means:** Unsanitized fields in approval prompts can inject content that tricks users into approving malicious actions.

**Remediation:**
```bash
# Update to v2026.2.14+ for the fix
openclaw update

# Always carefully read approval prompts before accepting
# Enable verbose approval mode
openclaw config set approvals.verbose true
```

### Check 45: Sandbox Browser Bridge Auth Bypass (CVE-2026-28468)

**Severity:** WARNING

**What it means:** The sandbox browser bridge server accepts requests without gateway authentication.

**Remediation:**
```bash
# Update to v2026.2.14+ for the fix
openclaw update

# Disable sandbox browser bridge if not needed
openclaw config set sandbox.browser.bridge false
```

### Check 46: Webhook DoS (CVE-2026-28478)

**Severity:** WARNING

**What it means:** Webhook handlers buffer request bodies without byte/time limits, allowing memory exhaustion via oversized or slow requests.

**Remediation:**
```bash
# Update to v2026.2.13+ for the fix
openclaw update

# Add reverse proxy with body size limits (nginx)
# client_max_body_size 1m;
```

### Check 47: TAR Archive Path Traversal (CVE-2026-28453)

**Severity:** CRITICAL

**What it means:** TAR archive extraction doesn't validate paths, allowing ../../ traversal to write files outside intended directories.

**Remediation:**
```bash
# Update to v2026.2.14+ for the fix
openclaw update

# Check for files written outside expected directories
find ~/.openclaw -name "*.tar" -o -name "*.tgz" -mtime -7 2>/dev/null
```

### Check 48: fetchWithGuard Memory DoS (CVE-2026-29609)

**Severity:** WARNING

**What it means:** fetchWithGuard allocates entire response payloads in memory before enforcing maxBytes, causing memory exhaustion.

**Remediation:**
```bash
# Update to v2026.2.14+ for the fix
openclaw update
```

### Check 49: /agent/act Auth Bypass (CVE-2026-28485)

**Severity:** CRITICAL

**What it means:** The /agent/act HTTP endpoint for browser control lacks authentication, allowing any local process to trigger browser actions.

**Remediation:**
```bash
# Update to v2026.2.12+ for the fix
openclaw update

# Disable browser extension if not needed
openclaw config set browser.extension.enabled false

# Check for unauthorized access attempts
grep "agent/act" ~/.openclaw/logs/*.log
```

### Check 50: Command Hijacking via PATH (CVE-2026-29610)

**Severity:** CRITICAL

**What it means:** OpenClaw resolves commands against PATH without pinning to absolute paths, allowing planted binaries to intercept exec calls.

**Remediation:**
```bash
# Update to v2026.2.14+ for the fix
openclaw update

# Remove writable directories before system dirs in PATH
# See Check 37 for detailed PATH hardening steps
```

### Check 51: SHA-1 Cache Poisoning (CVE-2026-28479, CVSS 8.7)

**Severity:** CRITICAL

**What it means:** SHA-1 is used for sandbox cache keys. Collision attacks can serve one sandbox's cached state to a different sandbox.

**Remediation:**
```bash
# Update to v2026.2.15+ for the fix
openclaw update

# Disable sandbox caching until upgraded
openclaw config set sandbox.cache.enabled false
```

### Check 52: Google Chat Webhook Bypass (CVE-2026-28469, CVSS 9.8)

**Severity:** CRITICAL

**What it means:** Google Chat webhook handler uses first-match semantics, allowing cross-account attackers to register matching paths and intercept/inject messages.

**Remediation:**
```bash
# Update to v2026.2.14+ for the fix
openclaw update

# Disable Google Chat integration if not actively used
openclaw config set integrations.googlechat.enabled false

# Audit webhook paths for duplicates
openclaw config get integrations.googlechat
```

### Check 53: Gateway WebSocket Device Identity Skip (CVE-2026-28472)

**Severity:** CRITICAL

**What it means:** Gateway WebSocket handshake skips device identity checks, granting operator access without device verification.

**Remediation:**
```bash
# Update to v2026.3.11+ for the fix
openclaw update

# Restrict gateway to localhost until upgraded
openclaw config set gateway.bind localhost

# Audit connected devices
openclaw device list
```

### Check 54: Cross-Site WebSocket Hijacking (CVE-2026-32302)

**Severity:** CRITICAL

**What it means:** Origin validation bypass in trusted-proxy mode allows attacker-origin pages to establish privileged operator WebSocket sessions.

**Remediation:**
```bash
# Update to v2026.3.11+ for the fix
openclaw update

# Disable trusted-proxy mode if not required
openclaw config set gateway.trustedProxy false

# Verify origin validation is active
curl -s -o /dev/null -w "%{http_code}" \
  -H "Origin: http://evil.attacker.com" \
  -H "Connection: Upgrade" -H "Upgrade: websocket" \
  http://127.0.0.1:18789/
# Should return 403, NOT 101
```

### Check 55: Device Pairing Credential Exposure (GHSA-7h7g)

**Severity:** WARNING

**What it means:** Device pairing setup codes expose long-lived gateway credentials instead of short-lived bootstrap tokens.

**Remediation:**
```bash
# Update to v2026.3.12+ for the fix
openclaw update

# Rotate device credentials after upgrading
openclaw device rotate-credentials

# Review paired devices
openclaw device list

# Revoke any suspicious device pairings
openclaw device revoke <device-id>
```

### Check 56: Operator Privilege Escalation (GHSA-vmhq)

**Severity:** CRITICAL

**What it means:** Accounts with operator.write permissions can access admin-only endpoints to create/delete browser profiles.

**Remediation:**
```bash
# Update to v2026.3.12+ for the fix
openclaw update

# Audit operator accounts
openclaw users list --role operator

# Remove unnecessary write permissions
openclaw users update <user> --role operator.read
```

### Check 57: MCP Server Tool Poisoning (OWASP MCP03/MCP06)

**Severity:** CRITICAL

**What it means:** MCP server configs contain hidden Unicode characters, prompt injection patterns, or BCC/forwarding directives used for tool poisoning attacks.

**Remediation:**
```bash
# Inspect MCP server configs for hidden content
cat -v ~/.openclaw/mcp-servers/*.json  # -v shows invisible chars

# Remove or quarantine suspicious configs
mv ~/.openclaw/mcp-servers/suspicious.json{,.quarantined}

# Use Cisco MCP Scanner for deeper analysis
# https://github.com/cisco-ai-defense/mcp-scanner

# Pin MCP server versions to prevent rug-pull updates
# Review tool descriptions after every MCP server update
```

### Check 58: SANDWORM_MODE MCP Worm Detection

**Severity:** CRITICAL

**What it means:** Your system contains artifacts from the SANDWORM_MODE npm worm — 19 typosquatted packages that inject rogue MCP servers into AI tool configs and harvest credentials.

**Remediation:**
```bash
# Remove rogue MCP entries from config files
# Check: ~/.claude.json, ~/.cursor/mcp.json, ~/.continue/config.json, ~/.windsurf/mcp.json

# Uninstall malicious npm packages
npm uninstall -g @anthropic/sdk-extra claude-code-utils claude-mcp-helper

# CRITICAL: Rotate ALL credentials
# - SSH keys: ssh-keygen (regenerate and update authorized_keys)
# - AWS keys: aws iam create-access-key / delete-access-key
# - npm tokens: npm token revoke / npm token create
# - LLM API keys: regenerate on each provider dashboard

# Audit git repos for unauthorized commits
git log --since="2 weeks ago" --oneline

# Check for unauthorized SSH keys
cat ~/.ssh/authorized_keys
```

### Check 59: Rules File Backdoor / Hidden Unicode Injection

**Severity:** CRITICAL

**What it means:** AI agent rules files (.cursorrules, CLAUDE.md, etc.) contain hidden Unicode characters (zero-width spaces, BiDi overrides) that inject invisible malicious instructions.

**Remediation:**
```bash
# Check for hidden Unicode in rules files
cat -v .cursorrules CLAUDE.md .clawrules 2>/dev/null | grep -P '[\x00-\x08\x0B\x0C\x0E-\x1F]'

# Strip hidden Unicode characters
perl -pi -e 's/[\x{200B}\x{200C}\x{200D}\x{2060}\x{FEFF}\x{00AD}]//g' .cursorrules
perl -pi -e 's/[\x{202A}-\x{202E}\x{2066}-\x{2069}]//g' CLAUDE.md

# Verify cleanup
hexdump -C .cursorrules | grep -E '200[bcd]|2060|feff|00ad'

# Set restrictive permissions on rules files
chmod 644 .cursorrules CLAUDE.md .clawrules 2>/dev/null
```

## Auto-Remediation

The `remediate.sh` orchestrator runs `scan.sh`, parses the results, skips CLEAN checks, and executes per-check remediation scripts for each WARNING/CRITICAL finding.

```bash
# Scan + remediate (interactive)
./scripts/remediate.sh

# Auto-approve all fixes (explicit opt-in)
OPENCLAW_ALLOW_UNATTENDED_REMEDIATE=1 ./scripts/remediate.sh --yes

# Dry run — preview what would be fixed
./scripts/remediate.sh --dry-run

# Remediate a single check
./scripts/remediate.sh --check 7 --dry-run

# Run all 41 scripts without scanning
./scripts/remediate.sh --all
```

### Per-Check Remediation Scripts

Each of the 41 scan checks has a dedicated remediation script in `scripts/remediate/`. Scripts are standalone, support `--yes` and `--dry-run`, and return exit 0 (fixed), 1 (failed), or 2 (nothing to fix). Unattended mode (`--yes`) requires `OPENCLAW_ALLOW_UNATTENDED_REMEDIATE=1`.

| Script | Check | Type |
|--------|-------|------|
| `check-01-c2-ips.sh` | C2 Infrastructure | Guidance |
| `check-02-malware-sigs.sh` | Malware Signatures & Obfuscation | Guidance |
| `check-03-reverse-shells.sh` | Reverse Shells | Auto-fix (Gatekeeper) |
| `check-04-exfil-endpoints.sh` | Credential Exfiltration | Auto-fix (/etc/hosts) |
| `check-05-crypto-wallet.sh` | Crypto Wallet Targeting | Guidance |
| `check-06-curl-pipe.sh` | Curl-Pipe Attacks | Auto-fix (tools.deny) |
| `check-07-file-perms.sh` | File & Credential Permission Audit | Auto-fix (chmod) |
| `check-08-skill-integrity.sh` | Skill Integrity | Guidance |
| `check-09-prompt-injection.sh` | AI Prompt Injection & Instruction Manipulation | Guidance |
| `check-10-gateway-config.sh` | Gateway Config | Auto-fix (bind/auth) |
| `check-11-websocket-sec.sh` | WebSocket Security | Guidance |
| `check-12-malicious-pub.sh` | Malicious Publishers | Guidance |
| `check-13-cred-leakage.sh` | Credential Leakage & Plaintext Secrets | Guidance |
| `check-14-dm-tool-sandbox.sh` | DM, Tool & Sandbox Policies | Auto-fix (config) |
| `check-15-mdns-exposure.sh` | mDNS/Bonjour Exposure | Auto-fix (disable) |
| `check-16-persistence.sh` | Persistence Mechanisms | Guidance |
| `check-17-log-security.sh` | Log Security & Poisoning | Auto-fix (sanitize) |
| `check-18-plugin-audit.sh` | Plugin/Extension Audit | Guidance |
| `check-19-docker-sec.sh` | Docker Security | Guidance |
| `check-20-auth-route-sec.sh` | Authentication & Route Security | Guidance |
| `check-21-exec-guardrails.sh` | Exec Guardrails & Approval Security | Guidance |
| `check-22-nodejs-cve.sh` | Node.js CVE Check | Guidance |
| `check-23-vscode-trojans.sh` | VS Code Trojans | Auto-fix (remove) |
| `check-24-internet-expose.sh` | Internet Exposure | Auto-fix (bind) |
| `check-25-mcp-security.sh` | MCP Server Security | Auto-fix (config) |
| `check-26-path-hijack.sh` | PATH Hijacking & Command Resolution | Auto-fix (remove) |
| `check-27-ssrf.sh` | SSRF Protection | Guidance |
| `check-28-path-traversal.sh` | Path Traversal & File Handling | Guidance |
| `check-29-dos-protection.sh` | DoS Protection | Guidance |
| `check-30-acp-approval.sh` | ACP Auto-Approval | Auto-fix (disable) |
| `check-31-env-override.sh` | Env Override Injection | Guidance |
| `check-32-privesc-scope.sh` | Privilege Escalation & Scope Abuse | Guidance |
| `check-33-sha1-cache.sh` | SHA-1 Cache Poisoning | Guidance |
| `check-34-gchat-webhook.sh` | Google Chat Webhook Bypass | Guidance |
| `check-35-sandworm.sh` | SANDWORM Worm Detection | Auto-fix (remove) |
| `check-36-workspace-plugin.sh` | Workspace Plugin Auto-Discovery | Guidance |
| `check-37-symlink-traversal.sh` | Symlink Traversal | Guidance |
| `check-38-sandbox-escape.sh` | Sandbox Escape & Session Inheritance | Guidance |
| `check-39-shell-env-rce.sh` | Shell Environment RCE | Guidance |
| `check-40-vnc-observer.sh` | VNC & Observer Authentication | Guidance |
| `check-41-device-identity-spoofing.sh` | Device Identity & Metadata Spoofing | Guidance |

**Auto-fix** scripts modify the system (permissions, config settings, /etc/hosts).
**Guidance** scripts print manual instructions for destructive or external actions (skill removal, credential rotation, Docker changes).

### Exit codes

| Code | Meaning |
|------|---------|
| 0 | Fixes applied successfully |
| 1 | Some fixes failed |
| 2 | Nothing to fix |

All actions are logged to `~/.openclaw/logs/remediation.log`.

## IOC Database

### C2 IP Addresses
| IP | Campaign | Notes |
|----|----------|-------|
| `91.92.242.30` | ClawHavoc | Primary AMOS C2, used by 335 skills |
| `95.92.242.30` | ClawHavoc | Secondary C2 |
| `96.92.242.30` | ClawHavoc | Secondary C2 |
| `54.91.154.110` | ClawHavoc | Reverse shell endpoint (port 13338) |
| `202.161.50.59` | ClawHavoc | Payload staging |

### Malicious Domains
| Domain | Type | Notes |
|--------|------|-------|
| `install.app-distribution.net` | Payload | AMOS installer distribution |
| `glot.io` | Hosting | Base64-obfuscated shell scripts (legitimate service abused) |
| `webhook.site` | Exfiltration | Data exfil via webhooks |
| `pipedream.net` | Exfiltration | Data exfil |
| `ngrok.io` | Tunneling | Reverse tunnel for exfiltration |
| `github.com/hedefbari` | Payload | Attacker GitHub hosting openclaw-agent.zip |
| `github.com/Ddoy233` | Payload | openclawcli repo - Windows infostealer |
| `download.setup-service.com` | Decoy | Decoy domain string in bash payload scripts |

### Known File Hashes (SHA-256)
| Hash | File | Platform |
|------|------|----------|
| `17703b3d...42283` | openclaw-agent.exe | Windows |
| `1e6d4b05...e2298` | x5ki60w1ih838sp7 | macOS (AMOS) |
| `0e52566c...4dd65` | unknown | macOS (AMOS variant) |
| `79e8f3f7...2bc1f2` | skill-archive | Any |

### Malicious Publisher Blacklist
| Publisher | Skills | Campaign | Source |
|-----------|--------|----------|--------|
| `hightower6eu` | 314 | ClawHavoc (crypto, finance, social lures) | Koi Security |
| `zaycv` | multiple | Bloom campaign (ClawHub + GitHub) | Bloom/JFrog |
| `noreplyboter` | 2 | Reverse shells (polymarket-all-in-one, better-polymarket) | Bloom/JFrog |
| `rjnpage` | 1 | .env exfiltration (rankaj) | Bloom/JFrog |
| `aslaep123` | multiple | Silent exfil (reddit-trends) | Bloom/JFrog |
| `gpaitai` | multiple | Bloom campaign | Bloom/JFrog |
| `lvy19811120-gif` | multiple | Bloom campaign | Bloom/JFrog |
| `Ddoy233` | 1 | Windows infostealer (openclawcli.zip) | Snyk/OpenSourceMalware |
| `hedefbari` | 1 | openclaw-agent.zip payload hosting | Koi Security |

### Skill Name Patterns
Malicious skills mimic popular categories:
- **Typosquats** (29): `clawhub`, `clawhubb`, `clawwhub`, `cllawhub`, `clawhubcli`, `clawdhub1`
- **Crypto** (111): `solana-wallet-*`, `phantom-wallet-*`, `bybit-agent`, `eth-gas-*`
- **Prediction markets** (34): `polymarket-*`, `better-polymarket`, `polymarket-all-in-one`
- **YouTube** (57): `youtube-summarize-*`, `youtube-*-pro`
- **Auto-updaters** (28): `auto-updat*`
- **Finance** (51): `yahoo-finance`, `stock-track*`
- **Google Workspace** (17): `google-workspace-*`, `gmail-*`, `gdrive-*`
- **Social/job lures** (Bloom): `reddit-trends`, `linkedin-job-application`, `reddit-*`, `linkedin-*`
- **Known malicious** (specific): `rankaj`, `openclawcli`

## IOC Auto-Updater

The `update-ioc.sh` script:
1. Checks upstream GitHub repo for IOC database updates
2. Scans active network connections against known C2 IPs
3. Validates installed skills against malicious publisher and pattern databases
4. Computes file hashes and checks against known malicious hashes
5. Rejects malformed IOC content and blocks untrusted feed override unless explicitly allowed

```bash
# Check for updates without applying
./scripts/update-ioc.sh --check-only

# Download and apply updates
./scripts/update-ioc.sh

# Override IOC source only with explicit trust bypass
OPENCLAW_ALLOW_UNTRUSTED_IOC_SOURCE=1 ./scripts/update-ioc.sh --github-repo https://example.com/ioc
```

## Web Dashboard

Zero-dependency Node.js server on port 18800 with:
- Scan summary donut chart
- 32 color-coded security check cards
- Gateway status and configuration audit
- Process tree via `witr` showing ancestry chains
- Network connections and listening ports
- File permissions and provider auth status
- Scan history timeline (last 30 scans)
- Auto-refresh every 30s + on-demand scan button

### Screenshots

**Dashboard overview** — header with scan summary donut, gateway status, and cron status:

![Dashboard Header](https://raw.githubusercontent.com/adibirzu/openclaw-security-monitor/main/docs/images/dashboard-header.png)

**Security checks grid** — all 32 checks with color-coded status cards (green=clean, yellow=warning, red=critical):

![Security Checks](https://raw.githubusercontent.com/adibirzu/openclaw-security-monitor/main/docs/images/security-checks.png)

**Full dashboard** — complete view including process tree, network connections, and scan history:

![Full Dashboard](https://raw.githubusercontent.com/adibirzu/openclaw-security-monitor/main/docs/images/dashboard-full.png)

**Process tree, network connections, and listening ports:**

![Process Tree & Network](https://raw.githubusercontent.com/adibirzu/openclaw-security-monitor/main/images/2.png)

**Provider auth, blocked commands, and scan history timeline:**

![Provider Auth & History](https://raw.githubusercontent.com/adibirzu/openclaw-security-monitor/main/images/3.png)

### API Endpoints

| Route | Method | Description |
|-------|--------|-------------|
| `/` | GET | Dashboard UI |
| `/api/scan` | GET | Latest scan results (parsed from log) |
| `/api/logs/scan` | GET | Scan log history (last 50 entries) |
| `/api/logs/cron` | GET | Cron log entries |
| `/api/ioc` | GET | IOC database stats (counts per file) |
| `/api/skills` | GET | Installed skills list with versions |
| `/api/config` | GET | Config audit (read-only) |
| `/api/status` | GET | Server uptime, last scan, IOC stats |
| `/api/help` | GET | CLI command reference |

### LaunchAgent (macOS)

The plist template in `docs/` uses `__HOME__` placeholders. Install with:

```bash
# Generate plist with your home directory and install
sed "s|__HOME__|$HOME|g" docs/com.openclaw.security-dashboard.plist \
  > ~/Library/LaunchAgents/com.openclaw.security-dashboard.plist

# Load the service
launchctl load ~/Library/LaunchAgents/com.openclaw.security-dashboard.plist

# Verify
curl -s http://localhost:18800/api/status | python3 -m json.tool
```

## CVE Coverage

| CVE | Description | Check |
|-----|-------------|-------|
| CVE-2026-25253 | 1-Click RCE via WebSocket hijacking | #14: WebSocket origin validation |
| CVE-2026-21636 | Permission model bypass (Node.js UDS sandbox escape) | #18: Tool policy audit, #19: Sandbox config, #28: Node.js version |
| CVE-2026-22708 | Indirect prompt injection | #9: SKILL.md injection, #10: Memory poisoning |
| CVE-2026-24763 | Command injection | #3: Reverse shell patterns |
| CVE-2026-25157 | Command injection | #6: Curl-pipe attacks |
| CVE-2026-24763 | Docker sandbox command injection | #27: Docker container security |
| CVE-2026-28363 | safeBins bypass via GNU long-option abbreviation (CVSS 9.9) | #35: safeBins validation |
| CVE-2026-28479 | SHA-1 sandbox cache key poisoning (CVSS 8.7) | #51: SHA-1 cache poisoning |
| CVE-2026-28485 | /agent/act unauthenticated access | #49: /agent/act auth |
| CVE-2026-29610 | Command hijacking via PATH | #50: PATH command hijacking |
| CVE-2026-28469 | Google Chat webhook cross-account bypass (CVSS 9.8) | #52: Google Chat webhook |
| CVE-2026-28472 | Gateway WebSocket device identity skip | #53: WebSocket device identity |
| CVE-2026-32302 | Cross-Site WebSocket Hijacking in trusted-proxy | #54: CSWSH protection |
| GHSA-7h7g | Device pairing credential exposure | #55: Pairing credentials |
| GHSA-vmhq | Operator privilege escalation | #56: Operator privesc |

## Threat Intelligence Sources

This project's detection patterns are built from published security research:

| Source | Report | What We Use |
|--------|--------|-------------|
| [Koi Security](https://www.koi.ai/blog/clawhavoc-341-malicious-clawedbot-skills-found-by-the-bot-they-were-targeting) | ClawHavoc: 341 Malicious Skills | C2 IPs, malicious publishers, skill patterns, AMOS indicators |
| [The Hacker News](https://thehackernews.com/2026/02/researchers-find-341-malicious-clawhub.html) | 341 Malicious ClawHub Skills | Campaign timeline, attack methodology |
| [VirusTotal](https://blog.virustotal.com/2026/02/from-automation-to-infection-how.html) | From Automation to Infection | File hashes, publisher identification, payload analysis |
| [Snyk](https://snyk.io/articles/skill-md-shell-access/) | From SKILL.md to Shell Access | SKILL.md injection patterns, memory poisoning techniques |
| [The Register](https://www.theregister.com/2026/02/02/openclaw_security_issues) | OpenClaw Security Issues | CVE details, ecosystem analysis |
| [SecurityWeek](https://www.securityweek.com/vulnerability-allows-hackers-to-hijack-openclaw-ai-assistant/) | Hijack OpenClaw AI Assistant | WebSocket hijacking details |
| [Cisco Blogs](https://blogs.cisco.com/ai/personal-ai-agents-like-openclaw-are-a-security-nightmare) | Personal AI Agents Security | Lethal trifecta analysis |
| [Tenable](https://www.tenable.com/blog/agentic-ai-security-how-to-mitigate-clawdbot-moltbot-openclaw-vulnerabilities) | Mitigate OpenClaw Vulnerabilities | Hardening recommendations |
| [1Password](https://1password.com/blog/from-magic-to-malware-how-openclaws-agent-skills-become-an-attack-surface) | From Magic to Malware | Skill attack surface analysis |
| [SOCRadar](https://socradar.io/blog/cve-2026-25253-rce-openclaw-auth-token/) | CVE-2026-25253 Analysis | Exploit chain technical details |
| [SOCPrime](https://socprime.com/active-threats/openclaw-ai-agent-weaponized/) | OpenClaw Weaponized | Detection and response patterns |
| [CyberInsider](https://cyberinsider.com/341-openclaw-skills-distribute-macos-malware-via-clickfix-instructions/) | ClickFix Malware Distribution | macOS-specific attack vectors |
| [eSecurity Planet](https://www.esecurityplanet.com/threats/hundreds-of-malicious-skills-found-in-openclaws-clawhub/) | Malicious Skills in ClawHub | Registry security analysis |
| [PointGuard AI](https://www.pointguardai.com/ai-security-incidents/openclaw-clawhub-malicious-skills-supply-chain-attack) | Supply Chain Attack Analysis | Attack timeline, scope |
| [SC Media](https://www.scworld.com/news/openclaw-agents-targeted-with-341-malicious-clawhub-skills) | 341 Malicious Skills | Technical indicators |
| [DepthFirst](https://depthfirst.com/post/1-click-rce-to-steal-your-moltbot-data-and-keys) | 1-Click RCE Exploit | CVE-2026-25253 original PoC research |
| [OpenClaw Security Docs](https://docs.openclaw.ai/gateway/security) | Official Security Guide | DM policies, sandbox config, tool restrictions, mDNS, log redaction |
| [DefectDojo](https://defectdojo.com/blog/the-openclaw-hardening-checklist-in-depth-edition) | Hardening Checklist (In-Depth) | File permissions, session isolation, incident response, audit commands |
| [Penligent AI](https://www.penligent.ai/hackinglabs/openclaw-sovereign-ai-security-manifest-a-comprehensive-post-mortem-and-architectural-hardening-guide-for-openclaw-ai-2026/) | Sovereign AI Security Manifest | Docker gVisor isolation, EASM, exploit path analysis, OIDC enforcement |
| [Vectra AI](https://www.vectra.ai/blog/clawdbot-to-moltbot-to-openclaw-when-automation-becomes-a-digital-backdoor) | When Automation Becomes a Backdoor | Host/network IoCs, persistence detection, credential harvesting patterns |
| [VentureBeat](https://venturebeat.com/security/openclaw-agentic-ai-security-risk-ciso-guide) | CISO Guide to Agentic AI Risk | Agentic hijacking patterns, blast radius analysis |
| [Guardz](https://guardz.com/blog/openclaw-hardening-for-msps/) | OpenClaw Hardening for MSPs | MSP-oriented hardening, kill switch procedures, skill verification |
| [Composio](https://composio.dev/blog/secure-openclaw-moltbot-clawdbot-setup) | Docker Hardening Guide | Non-root containers, read-only FS, restricted outbound, capability dropping |
| [AuthMind](https://www.authmind.com/post/openclaw-malicious-skills-agentic-ai-supply-chain) | Agentic AI Supply Chain Attack | Identity security, agent capability dependencies |
| [BleepingComputer](https://www.bleepingcomputer.com/news/security/malicious-moltbot-skills-used-to-push-password-stealing-malware/) | MoltBot Malware Distribution | 400+ malware packages timeline, password stealer delivery |
| [SecurityAffairs](https://securityaffairs.com/187562/malware/moltbot-skills-exploited-to-distribute-400-malware-packages-in-days.html) | 400+ Malware Packages in Days | MoltBot/OpenClaw exploitation timeline and scale |
| [TechBuzz](https://www.techbuzz.ai/articles/openclaw-ai-agent-goes-viral-despite-security-flaws) | OpenClaw Goes Viral Despite Flaws | Ecosystem analysis, security tradeoffs |
| [Consortium.net](https://consortium.net/blog/security-advisory-openclaw-moltbot-ai-agents) | Security Advisory | Comprehensive advisory for OpenClaw/Moltbot deployments |
| [Security.com](https://www.security.com/expert-perspectives/rise-openclaw) | The Rise of OpenClaw | Expert perspectives on OpenClaw security landscape |
| [OpenSourceMalware](https://opensourcemalware.com/blog/clawdbot-skills-ganked-your-crypto) | ClawdBot Skills Ganked Your Crypto | Original ClawHavoc disclosure, zaycv/Ddoy233 accounts, payload analysis |
| [Bloom Security / JFrog](https://jfrog.com/blog/giving-openclaw-the-keys-to-your-kingdom-read-this-first/) | Keys to Your Kingdom | 37 malicious skills, 3 distinct campaigns, 10 attacker GitHub accounts |
| [Infosecurity Magazine](https://www.infosecurity-magazine.com/news/malicious-crypto-trading-skills/) | Malicious Crypto Trading Skills | ByBit/Polymarket/Axiom/Reddit/LinkedIn targeting |
| [OODA Loop](https://oodaloop.com/briefs/technology/malicious-openclaw-skill-targets-crypto-users-on-clawhub/) | Malicious Skill Targets Crypto Users | ClawHub marketplace attack analysis |
| [ThaiCERT](https://www.thaicert.or.th/en/2026/02/04/researchers-discover-over-200-malicious-skills-on-openclaw-distributing-password-stealing-malware/) | 200+ Malicious Skills Advisory | National CERT advisory on OpenClaw malware distribution |
| [Snyk](https://snyk.io/articles/clawdhub-malicious-campaign-ai-agent-skills/) | clawdhub Campaign Deep-Dive | Base64 payload chains, glot.io hosting, zaycv/Ddoy233 accounts, NovaStealer |
| [CSA AI Safety Initiative](https://cloudsecurityalliance.org) | Moltbook and OpenClaw Security Research Note | Moltbook DB exposure (1.5M API tokens), agent impersonation, lethal trifecta, enterprise risk assessment |
| [Wiz Research](https://www.wiz.io/blog/exposed-moltbook-database-reveals-millions-of-api-keys) | Hacking Moltbook: 1.5M API Keys | Moltbook Supabase exposure, plaintext OpenAI keys, agent credential theft |
| [Palo Alto Networks](https://venturebeat.com/security/openclaw-agentic-ai-security-risk-ciso-guide) | Lethal Trifecta Analysis | Private data access + untrusted content + external communication |
| [CrowdStrike](https://www.crowdstrike.com/en-us/blog/what-security-teams-need-to-know-about-openclaw-ai-super-agent/) | What Security Teams Need to Know | Enterprise detection, Falcon for IT, process tree analysis |
| [OWASP GenAI](https://genai.owasp.org/resource/owasp-top-10-for-agentic-applications-for-2026/) | Top 10 for Agentic Applications 2026 | Agent goal hijack, tool misuse, identity abuse, MCP risks |
| [ToxSec](https://www.toxsec.com/p/openclaw-security-checklist) | OpenClaw Security Checklist | Command allowlisting, Docker hardening, MCP audit, egress filtering |
| [Brandefense](https://brandefense.io/blog/unmanaged-shadow-ai-agent/) | Unmanaged Shadow AI Agent | Shodan exposure (21K+ instances), internet exposure detection |
| [Aikido.dev](https://www.aikido.dev/blog/fake-clawdbot-vscode-extension-malware) | Fake ClawdBot VS Code Extension | ScreenConnect RAT, VS Code trojan detection |
| [Permiso](https://permiso.io/blog/inside-the-openclaw-ecosystem-ai-agents-with-privileged-credentials) | AI Agents with Privileged Credentials | Identity abuse, credential chain analysis |
| [Argus Security Audit](https://github.com/openclaw/openclaw/issues/1796) | 512 Audit Findings | CSRF, OAuth bypass, plaintext credentials, exec-approvals |
| [yevhen/clawdbot-security-scanner](https://github.com/yevhen/clawdbot-security-scanner) | 37-Check Security Scanner | Token reuse, Tailscale bypass, browser control, session scope |
| [Prompt Security](https://prompt.security/blog/top-10-mcp-security-risks) | Top 10 MCP Security Risks | Tool poisoning, rug pulls, command injection via MCP |
| [Hudson Rock](https://www.infostealers.com/article/hudson-rock-identifies-real-world-infostealer-infection-targeting-openclaw-configurations/) | Vidar Targets Agent Identity (Feb 13) | First infostealer targeting OpenClaw config dirs (soul.md, device.json) |
| [Eye Security](https://research.eye.security/log-poisoning-in-openclaw/) | Log Poisoning in OpenClaw (Feb 2026) | WebSocket header injection into agent logs, indirect prompt injection |
| [Snyk](https://snyk.io/blog/toxicskills-malicious-ai-agent-skills-clawhub/) | ToxicSkills: 36% of Skills Flawed (Feb 5) | 3,984 skills scanned, 534 critical, 76 confirmed malicious payloads |
| [Bitdefender](https://businessinsights.bitdefender.com/technical-advisory-openclaw-exploitation-enterprise-networks) | Technical Advisory: Enterprise Exploitation | 14 malicious actors, GravityZone telemetry, Shadow AI in enterprises |
| [SecurityScorecard](https://securityscorecard.com) | STRIKE: 135K Exposed Instances (Feb 9) | 135,000+ instances, 12,812 RCE-exploitable, 1,800 leaking keys |
| [Antiy CERT](https://gbhackers.com/clawhavoc-infects-openclaws-clawhub/) | ClawHavoc: 1,184 Packages, 12 Accounts | Expanded campaign analysis, automated submission patterns |
| [Endor Labs](https://www.endorlabs.com/learn/how-ai-sast-traced-data-flows-to-uncover-six-openclaw-vulnerabilities) | AI SAST: 6 Vulnerabilities with PoC | SSRF, missing auth, path traversal discovered via AI SAST |
| [Adversa AI](https://adversa.ai/blog/secureclaw-open-source-ai-agent-security-for-openclaw-aligned-with-owasp-mitre-frameworks/) | SecureClaw: OWASP-Aligned Security (Feb 18) | New open-source security plugin aligned with OWASP ASI Top 10 |
| [Kaspersky](https://kaspersky.com) | OpenClaw Risks Analysis (Feb 2026) | Clawdbot/Moltbot key risks for enterprise deployments |
| [Palo Alto Networks](https://www.paloaltonetworks.com/blog/network-security/why-moltbot-may-signal-ai-crisis/) | Next AI Security Crisis (Feb 2026) | Enterprise security team guidance |
| [Trend Micro](https://trendmicro.com) | Agentic Assistants Analysis (Feb 2026) | What OpenClaw reveals about agentic assistant security |
| [Oasis Security](https://www.oasis.security/blog/openclaw-vulnerability) | ClawJacked: WebSocket Brute-Force (Feb 26) | Localhost WebSocket hijack via password brute-force, auto-device registration |
| [Flare](https://flare.io/learn/resources/blog/widespread-openclaw-exploitation) | Widespread Exploitation by Multiple Groups (Feb 25) | Multiple threat groups targeting OpenClaw in the wild |
| [CVE-2026-28363](https://advisories.gitlab.com/pkg/npm/openclaw/CVE-2026-28363/) | safeBins Bypass via sort --compress-prog (CVSS 9.9) | GNU long-option abbreviation bypasses exec allowlist |
| [GHSA-7jx5](https://advisories.gitlab.com/pkg/npm/openclaw/GHSA-7jx5-9fjg-hp4m/) | ACP Auto-Approval Bypass (CVSS 8.2) | Untrusted metadata bypasses interactive tool approval |
| [CVE-2026-28446](https://dev.to/tiamatenity/cve-2026-28446-cvss-98-openclaw-voice-extension-rce-what-you-need-to-know-n2i) | Voice-Call Extension RCE (CVSS 9.8) | Pre-auth RCE via crafted audio payload, 42K instances exploitable |
| [CVE-2026-28484](https://www.sentinelone.com/vulnerability-database/cve-2026-28484/) | Git Pre-Commit Hook RCE (CVSS 9.3) | Command injection via crafted filenames in staged files |
| [CVE-2026-28458](https://www.redpacketsecurity.com/cve-alert-cve-2026-28458-openclaw-openclaw/) | Browser Relay CDP Auth Bypass (CVSS 7.5) | Unauthenticated /cdp WebSocket enables session theft |
| [CVE-2026-28462](https://www.redpacketsecurity.com/cve-alert-cve-2026-28462-openclaw-openclaw/) | Browser Control Path Traversal (CVSS 7.5) | Arbitrary file writes via /trace/stop, /download |
| [CVE-2026-29609](https://www.redpacketsecurity.com/cve-alert-cve-2026-29609-openclaw-openclaw/) | fetchWithGuard Memory DoS (CVSS 7.5) | Memory exhaustion via oversized responses |
| [Huntress](https://www.huntress.com/blog/openclaw-github-ghostsocks-infostealer) | Fake OpenClaw Installers (Mar 4) | GhostSocks + Vidar via Stealth Packer, Bing AI search poisoning |
| [BleepingComputer](https://www.bleepingcomputer.com/news/security/bing-ai-promoted-fake-openclaw-github-repo-pushing-info-stealing-malware/) | Bing AI Promoted Fake Installers | Search engine poisoning distributing malware via GitHub |
| [MintMCP](https://www.mintmcp.com/blog/openclaw-cve-explained) | Every OpenClaw CVE Explained | Comprehensive CVE reference for enterprise teams |
| [jgamblin/OpenClawCVEs](https://github.com/jgamblin/OpenClawCVEs/) | OpenClaw CVE Tracker | Community-maintained CVE tracking repository |
| [Check Point](https://research.checkpoint.com/2026/rce-and-api-token-exfiltration-through-claude-code-project-files-cve-2025-59536/) | Claude Code MCP Flaws | RCE and API key exfiltration via MCP/hooks config |
| [Socket](https://socket.dev/blog/sandworm-mode-npm-worm-ai-toolchain-poisoning) | SANDWORM_MODE npm Worm (Feb 20) | 19 typosquatted npm packages carrying MCP worm malware |
| [Pillar Security](https://www.pillar.security/blog/new-vulnerability-in-github-copilot-and-cursor-how-hackers-can-weaponize-code-agents) | Rules File Backdoor | Hidden Unicode in AI rules files injects invisible instructions |
| [OWASP](https://owasp.org/www-project-mcp-top-10/) | MCP Top 10 (2026) | Tool poisoning, supply chain, command injection via MCP |
| [CyberArk](https://www.cyberark.com/resources/threat-research-blog/poison-everywhere-no-output-from-your-mcp-server-is-safe) | MCP Output Poisoning | Full-schema poisoning via function names, defaults, required fields |
| [Semgrep](https://semgrep.dev/blog/2025/so-the-first-malicious-mcp-server-has-been-found-on-npm-what-does-this-mean-for-mcp-security/) | First Malicious MCP Server | postmark-mcp rug pull: BCC exfiltration after 15 clean versions |
| [Unit42/Palo Alto](https://unit42.paloaltonetworks.com/model-context-protocol-attack-vectors/) | MCP Sampling Attack Vectors | Resource theft, conversation hijacking, covert tool invocation |
| [LayerX](https://layerxsecurity.com/blog/claude-desktop-extensions-rce/) | Claude DXT Zero-Click RCE | Calendar event injection into unsandboxed DXT extensions |
| [CVE-2026-28469](https://dailycve.com/openclaw-authorization-bypass-cve-2026-28469-critical/) | Google Chat Webhook Bypass (CVSS 9.8) | Cross-account authorization bypass via first-match semantics |
| [CVE-2026-32302](https://cvereports.com/reports/CVE-2026-32302) | CSWSH in Trusted-Proxy Mode | Origin validation bypass allows attacker WebSocket sessions |
| [GHSA-vmhq](https://cvereports.com/reports/GHSA-VMHQ-CQM9-6P7Q) | Operator Privilege Escalation | operator.write accounts access admin endpoints |

## Related Security Tools

Other open-source projects for securing OpenClaw deployments:

| Tool | Description | URL |
|------|-------------|-----|
| **Cisco Skill Scanner** | Static + behavioral + LLM analysis of agent skills (YARA, AST, VirusTotal) | [cisco-ai-defense/skill-scanner](https://github.com/cisco-ai-defense/skill-scanner) |
| **Cisco MCP Scanner** | Scan MCP servers for threats and security findings | [cisco-ai-defense/mcp-scanner](https://github.com/cisco-ai-defense/mcp-scanner) |
| **openclaw-secure-start** | Pre-first-run hardening: gateway binding, tool policies, VPS/Pi configs | [pottertech/openclaw-secure-start](https://github.com/pottertech/openclaw-secure-start) |
| **Clawdex** | Koi Security's skill reputation checker (web + installed skill scan) | [koi.ai](https://www.koi.ai) |
| **skillvet** | Security scanner for ClawHub skills: malware, exfil, prompt injection | [ClawHub](https://github.com/openclaw/skills) |
| **skill-flag** | Pattern-based backdoor and malware scanner for OpenClaw skills | [ClawHub](https://github.com/openclaw/skills) |
| **clawdefender** | Security scanner and input sanitizer for AI agents | [ClawHub](https://github.com/openclaw/skills) |
| **agentguard** | Security monitoring skill for OpenClaw agents | [ClawHub](https://github.com/openclaw/skills) |
| **prompt-guard** | Prompt injection defense system for OpenClaw/Clawdbot | [ClawHub](https://github.com/openclaw/skills) |

**How this project differs:** openclaw-security-monitor runs *outside* the agent as independent host-level monitoring — it scans the filesystem, network, processes, and gateway configuration directly, without relying on the agent's own tools or trust boundaries. This provides defense-in-depth even if the agent itself is compromised.

## Install as OpenClaw Skill

Deploy directly into the OpenClaw agent's skill directory for in-agent access:

```bash
# From GitHub
git clone https://github.com/adibirzu/openclaw-security-monitor.git \
  ~/.openclaw/workspace/skills/<skill-dir>
chmod +x ~/.openclaw/workspace/skills/<skill-dir>/scripts/*.sh

Replace `<skill-dir>` with the actual folder name where the skill is installed (commonly `openclaw-security-monitor` or `security-monitor`).
```

The agent auto-discovers skills from `~/.openclaw/workspace/skills/` via SKILL.md frontmatter. After installation, these commands are available in the agent:

| Command | Description |
|---------|-------------|
| `/security-scan` | Run 41-point security scan |
| `/security-remediate` | Auto-fix common findings |
| `/security-dashboard` | CLI security dashboard |
| `/security-network` | Network connection monitor |
| `/security-setup-telegram` | Configure Telegram alerts |

To verify installation:
```bash
# Check skill is visible
openclaw skills list | grep openclaw-security-monitor

# Test scan through the agent
# In the agent, type: /security-scan
```

## Requirements

- macOS or Linux
- bash 4+
- Node.js 22.12+ (for web dashboard; older versions are flagged by the scanner)
- `curl` (for IOC updates and WebSocket check)
- Optional: [`witr`](https://github.com/pranshuparmar/witr) for process tree analysis
- Optional: `openclaw` for gateway configuration audit

### Install Prerequisites (macOS)

```bash
# Node.js (required for web dashboard)
brew install node@22

# witr - process ancestry tracer (optional but recommended)
brew install witr

# Verify
node --version   # v22.x
witr --version   # should print version
```

### Install Prerequisites (Linux)

```bash
# Node.js (via NodeSource or your package manager)
curl -fsSL https://deb.nodesource.com/setup_22.x | sudo bash -
sudo apt-get install -y nodejs

# witr (check https://github.com/pranshuparmar/witr for Linux install)
```

### Environment Variables

| Variable | Required | Description |
|----------|----------|-------------|
| `OPENCLAW_HOME` | No | OpenClaw home directory (default: `~/.openclaw`) |
| `OPENCLAW_TELEGRAM_TOKEN` | For alerts | Telegram bot token for scan notifications |
| `DASHBOARD_HOST` | No | Dashboard bind address (default: `127.0.0.1`, use `0.0.0.0` for LAN access) |

## Contributing

IOC updates are welcome. To add new indicators:

1. Fork the repo
2. Add entries to the appropriate `ioc/*.txt` file following the existing format
3. Submit a PR with the source reference

## License

MIT

openclaw-security-monitor

openclaw-security-monitor 최신버젼 다운로드

오늘의 추천 프로젝트

Was this content helpful?

Recent Feedback

openclaw-security-monitor

openclaw-security-monitor 최신버젼 다운로드

오늘의 추천 프로젝트

Was this content helpful?

Recent Feedback

Download README