Skip to content

Deploy an Agent Workflow in 5 Minutes

Deploy a visual AI agent builder with compliance auditing. This guide uses Dify with Lucid's workflow profile for prompt injection protection and SOC2 compliance controls.

Alpha Access Required

Lucid is in private alpha. Request access before proceeding.

Time to complete: ~4 minutes

What You'll Get

At the end of this quickstart, you'll have:

  • A live Dify agent builder at a unique URL
  • Llama 3.1 70B model for complex reasoning tasks
  • Compliance auditors protecting every workflow:
  • Prompt injection detection
  • SOC2 controls (PII detection, credential scanning)
  • AI Passports for audit trails and compliance reporting

Step 1: Install and Authenticate

If you haven't already, install the CLI and log in:

# Install the CLI
pip install lucid-cli

# Log in to your account
lucid login -e your@email.com -p yourpassword

Expected output:

Logged in as your@email.com

Step 2: Deploy Your Agent Builder

Run a single command to deploy:

lucid apply --app dify --model llama-3.1-70b --profile workflow

Expected output:

[*] Creating serverless environment...
[+] Environment created: env-xyz789ghi012

    Connection URL: https://env-xyz789ghi012.serverless.lucid.ai
    App:            dify
    Model:          meta-llama/Llama-3.1-70B-Instruct
    Auditors:       injection, soc2
    Region:         us-east-1

[+] Environment ready!

Step 3: Access Your Agent Builder

Open the Connection URL in your browser:

https://env-xyz789ghi012.serverless.lucid.ai

You now have access to Dify's visual agent builder. Try creating your first agent:

  1. Click "Create App" and select "Agent"
  2. Name it "Customer Support Bot"
  3. Add a system prompt: "You are a helpful customer support agent."
  4. Test it in the preview panel

Every interaction is automatically audited for security and compliance.

Step 4: Verify Security (Optional)

Confirm your environment has valid TEE attestation:

lucid verify environment env-xyz789ghi012

Expected output:

[*] Fetching routing info for environment env-xyz789ghi012...
  [+] Model: https://model-abc.serverless.lucid.ai (us-east-1) - amd_sev_snp
  [+] Auditor: https://auditor-def.serverless.lucid.ai (us-east-1) - amd_sev_snp
  [+] App: https://app-ghi.serverless.lucid.ai (us-east-1) - amd_sev_snp

[*] 3/3 endpoints have attestation reports

What the Workflow Profile Includes

The workflow profile activates these auditors:

Auditor What It Does
Injection Blocks prompt injection and jailbreak attempts
SOC2 PII detection, credential scanning, access logging

SOC2 Compliance Features

The SOC2 auditor provides:

  • PII Detection: Identifies personal data (names, emails, SSNs, etc.)
  • Credential Scanning: Blocks API keys, passwords, tokens from leaking
  • Access Logging: Records all interactions for audit trails
  • Data Classification: Tags sensitive data for compliance reporting

Alternative: Deploy n8n

Prefer n8n for workflow automation? Just change the app:

lucid apply --app n8n --model llama-3.1-70b --profile workflow

n8n provides:

  • 400+ integrations (Slack, GitHub, databases, etc.)
  • Visual workflow designer
  • AI nodes for LLM-powered automation

Build a Sample Agent Workflow

Once in Dify, try this example workflow:

Customer Feedback Analyzer

  1. Create a new Workflow app
  2. Add an LLM node with this prompt:
    Analyze the following customer feedback and extract:
    1. Sentiment (positive/negative/neutral)
    2. Key topics mentioned
    3. Suggested action items
    
    Feedback: {{input}}
    
  3. Connect to an output node
  4. Test with sample feedback:
    "The product is great but shipping took too long.
    Would buy again if delivery improves."
    

The workflow runs through the auditor chain, ensuring no PII leaks and no injection attempts succeed.

View Compliance Reports

Check your AI Passports for compliance documentation:

lucid passport list
ID              AGENT           TIMESTAMP
pass-101        env-xyz789      2024-01-15T14:00:00Z
pass-102        env-xyz789      2024-01-15T14:05:00Z

View detailed compliance data:

lucid passport show pass-101
Passport ID: pass-101
Hardware Attested: true
TEE Type: AMD SEV-SNP
Auditors: injection, soc2
PII Detected: false
Credentials Detected: false
Injection Blocked: false

EU Data Residency

For GDPR compliance, deploy to the EU region:

lucid apply --app dify --model llama-3.1-70b --profile workflow --region eu

Clean Up

When you're done experimenting, you can manage your environment through the Observer dashboard at observer.lucid.sh or use the status command:

lucid status

Serverless environments can be stopped via the Observer UI. For local development environments, use:

lucid teardown

Next Steps

Summary

You deployed a compliant agent builder with:

Component Value
App Dify
Model Llama 3.1 70B
Auditors injection, soc2
TEE AMD SEV-SNP
Time ~4 minutes

Your agent workflows are now protected by hardware-backed security with compliance controls meeting SOC2 requirements.