SudoDog - Monitor All Your AI Agents In One Place

Monitor all your AI agents in one place

See what every agent is doing. Real-time visibility, complete audit trail, resource monitoring. Works with LangChain, AutoGPT, CrewAI, or any framework. No code changes—just prepend sudodog run.

Start Monitoring Free See It In Action

📖 Read the 5-Minute Tutorial • 📚 Full Documentation

bash • quick install 
$ curl -sL https://sudodog.com/install.sh | bash
● One command · zero configuration

See It In Action

Zero integration - just prepend sudodog run to your existing command

bash • sudodog run --docker --image my-agent:latest python agent.py 
user@prod:~$ sudodog run --docker --image my-agent:latest python agent.py
🐕 SudoDog AI Agent Security
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
🐳 Using Docker sandbox
🐳 Creating Docker container...
   Image: my-agent:latest
   Network: enabled
   CPU limit: 2.0 cores
   Memory limit: 1g
✓ Container created: d2dcfb3e93d6
▶ Starting container d2dcfb3e93d6...
✓ Container running

🤖 AI Agent Starting...
[Safe Operation] Reading OS information...
[Safe Operation] Creating temporary file...

[Dangerous Operation] Attempting to read /etc/shadow...
  ✓ Blocked by container isolation

[Dangerous Operation] Simulating database query...
  Query: DROP TABLE customers;
  ⚠️ Pattern detected by SudoDog!

✓ Container exited with code 0
CPU: 15.3% | Memory: 245.8MB
⏹ Stopping container d2dcfb3e93d6...
✓ Container removed

user@prod:~$ sudodog daemon status
✓ Daemon is running (PID: 95100)
Active containers: 3 | CPU avg: 12.5% | Memory avg: 28.3%

Complete Visibility & Control For AI Agents

👁️

Real-Time Monitoring

See all agents at once. Live CPU and memory stats. Background daemon tracks every container. Know exactly what each agent is doing right now.

📊

Complete Audit Trail

Immutable logs of every agent action in JSONL format. Track all operations with timestamps. Perfect for debugging and compliance.

🛡️

Pattern Detection

Detects dangerous operations like SQL injection (DROP TABLE) and shell commands (rm -rf, curl | bash). Know when agents attempt risky actions.

🚀

Framework-Agnostic

Works with LangChain, AutoGPT, CrewAI, or any custom agent. Node.js, Python, Ruby—any language. No code changes needed.

💪

Resource Monitoring

Track CPU and memory usage per agent. Set resource limits to prevent runaway processes. Get alerts when thresholds are exceeded.

🐳

Docker Isolation

Run agents in isolated containers with custom images. Full dependency support. Strong isolation for production deployments.

From CLI To Web Dashboard

Start monitoring locally, upgrade to web dashboard when you need team visibility

Free

Available Now

Monitor agents locally with CLI dashboard

  • Unlimited agents
  • CLI dashboard - real-time monitoring
  • Framework-agnostic - any agent, any language
  • Docker isolation with custom images
  • Resource monitoring - CPU & memory tracking
  • Pattern detection (SQL, shell)
  • Local audit logs (JSONL)
  • File rollback
  • Sample agents included
  • Open source (MIT)
  • Community support
Start Monitoring Free
Coming Soon

Pro

Join Waitlist

Web dashboard for teams monitoring agents

  • Everything in Free
  • Web dashboard - see all agents, all servers
  • Anomaly detection
  • Threat intelligence feed - emerging patterns
  • Email/Slack alerts - instant notifications
  • Cloud logging - searchable history
  • Historical analytics - spot trends over time
  • Team collaboration - up to 5 users
  • Multi-server support
  • 30-day log retention
  • Priority support
Join Waitlist
Coming Soon

Enterprise

Join Waitlist

Advanced compliance and security for large teams

  • Everything in Pro
  • Unlimited team members
  • SOC 2 compliance reports - auto-generated
  • Custom data retention - 1 year+
  • SSO integration (SAML, OAuth)
  • SIEM integration (Splunk, Datadog)
  • On-premise deployment option
  • Custom integrations & API
  • SLA & premium support
Join Waitlist

Zero Integration - Just Prepend One Command

Already have an AI agent? Start monitoring it instantly with no code changes.

# Before: Your existing agent
python my_langchain_agent.py
node my_autogpt_agent.js
ruby my_agent.rb

# After: Secured with SudoDog (that's it!)
sudodog run python my_langchain_agent.py
sudodog run node my_autogpt_agent.js
sudodog run ruby my_agent.rb

Installation & First Run

Get started in 30 seconds with sample agents included.

# Step 1: Install SudoDog (creates sample agents in ~/sudodog-examples)
curl -sL https://sudodog.com/install.sh | bash

# Step 2: Initialize SudoDog
sudodog init

# Step 3: Test with sample agent to verify it works
sudodog run python ~/sudodog-examples/hello_agent.py

# Step 4: Now run YOUR OWN agent!
sudodog run python your_agent.py

# Step 5: View the security audit trail
sudodog logs

Real Framework Examples

Works with any AI framework out of the box

# LangChain
sudodog run python langchain_agent.py

# AutoGPT
sudodog run python -m autogpt

# CrewAI
sudodog run python crewai_agent.py

# Custom Python agent
sudodog run python custom_agent.py

# Any command works!
sudodog run node agent.js
sudodog run ./agent.sh

Production: Docker + Real-time Monitoring

For production-grade isolation, install Docker and unlock all features

# Install Docker (one-time setup)
curl -fsSL https://get.docker.com | sudo sh
sudo usermod -aG docker $USER
# Log out and back in for docker group to take effect

# Run YOUR agent with Docker isolation
sudodog run --docker python your_agent.py

# Start real-time monitoring daemon
sudodog daemon start

# Run with custom image and resource limits
sudodog run --docker --image my-agent:latest \
  --cpu-limit 2.0 --memory-limit 1g python your_agent.py

# Check container stats
sudodog daemon status

Works on Ubuntu, Debian, Arch, Fedora, and most Linux distributions.

📚 Full Documentation: Quick Start Guide • 5-Minute Tutorial

Open Source: View on GitHub • MIT License