Skip to main content
This page provides practical examples of using the Meta Agent for common tasks.

Example 1: CVE Remediation Workflow

This walkthrough demonstrates the Meta Agent’s end-to-end capabilities by scanning a repository for CVEs and automatically creating a pull request with fixes.

Step 1: Enable Connectors

First, enable the connectors you’ll need. Click the connectors menu and enable GitHub and Sandbox: Enabling Connectors

Step 2: Request CVE Scan and Remediation

Enter a prompt asking the Meta Agent to scan for vulnerabilities and fix them:
Scan the repository kubiyabot/example-app for CVEs and create
a pull request with fixes for any critical vulnerabilities found.

Step 3: Review Results

The Meta Agent will:
  1. Clone the repository in a sandbox environment
  2. Run vulnerability scanning tools (like Trivy)
  3. Analyze the results and identify fixable issues
  4. Generate updated dependency files
  5. Create a pull request with the remediation

What You Get

  • Scan Results - List of vulnerabilities with severity levels
  • Remediation Plan - What will be changed and why
  • Pull Request - Ready for review with detailed description
  • Slack Notification - Optional alert to your team

Example 2: Infrastructure Discovery

Query your infrastructure without writing Cypher:

Find Dependencies

What services depend on the user-auth database?
Response includes:
  • List of dependent services
  • Relationship types (CONNECTS_TO, DEPENDS_ON)
  • Impact assessment for changes

Permission Analysis

What IAM roles have administrator access across our AWS accounts?
Response includes:
  • Role names and ARNs
  • Attached policies
  • Account breakdown
  • Security recommendations

Resource Discovery

List all EC2 instances in production with their IAM roles
Response includes:
  • Instance IDs and names
  • Associated IAM roles
  • Region and account information
  • Running status

Example 3: Agent Execution

This example demonstrates executing an agent on a remote worker through Meta Agent’s durable task management.

List and Execute Agents

Ask the Meta Agent to list available agents and execute one:
List all available agents in the platform, then execute the DevOps
agent to check the status of our Kubernetes infrastructure
Meta Agent Lists Agents The Meta Agent will:
  1. List all configured agents with their capabilities
  2. Identify available worker queues with active workers
  3. Create a durable task and route it to the appropriate queue
  4. Stream real-time execution output as the agent works

Live Task Execution

Watch the agent execute in real-time with full visibility into tool calls: Live Task Execution What you see:
  • Task Started indicator with execution ID
  • Runtime (Claude Code) and Agent (DevOps Engineer) details
  • Worker Queue where the task is executing
  • Live message stream showing agent reasoning and tool calls
  • Tool execution status (Running, Success, Failed)

Create a New Agent

Create a new agent called "Database Admin" with PostgreSQL and
MySQL skills, assign it to the Platform team
Meta Agent will:
  1. Create the agent with specified configuration
  2. Attach the requested skills
  3. Assign to the team
  4. Return the agent details

Example 4: Team Coordination

Create a Team

Create a Platform Engineering team with the DevOps agent and
SRE agent, give them access to production environment

Check Team Status

Show me all teams and which agents are assigned to each

Execute Team Workflow

Have the Security team run a full compliance audit on our
AWS infrastructure

Example 5: Memory and Learning

Query Past Incidents

What did we learn from the last production outage?
Response includes:
  • Incident summary from cognitive memory
  • Root cause analysis
  • Remediation steps taken
  • Lessons learned

Store New Knowledge

Store this runbook for future reference: When the API gateway
returns 503 errors, first check the upstream services...

Find Runbooks

Find the runbook for database failover procedures

Example 6: Background Jobs

Schedule a Job

Schedule a daily backup verification job that runs at 2am UTC

Check Job Status

Show me all background jobs and their last execution status

Pause/Resume Jobs

Pause the nightly report job until next Monday

Example 7: Cross-Platform Operations

Create Project with Full Setup

Create a new project for Q1 cost optimization:
- Assign the FinOps team
- Schedule weekly cloud spend analysis
- Create a Slack channel for updates
- Set up a Jira board for tracking
Meta Agent orchestrates:
  1. Project creation
  2. Team assignment
  3. Background job scheduling
  4. Slack channel creation (with Slack connector)
  5. Jira board setup (with Jira connector)

Best Practices

  1. Be Specific - Include context about what you’re looking for
  2. Leverage Natural Language - Write as you would explain to a colleague
  3. Follow Up - Ask clarifying questions based on responses
  4. Use Suggestions - Act on AI-powered insights for proactive management
  5. Review Tool Calls - Check which platform resources were accessed
  6. Enable Appropriate Connectors - Make sure needed integrations are enabled

What’s Next