Skip to content

sample_tasks.md

Latest commit

 

History

History
344 lines (242 loc) · 8.34 KB

sample_tasks.md

File metadata and controls

344 lines (242 loc) · 8.34 KB

Sample Tasks and Usage Examples

Basic Usage

Example 1: Create a REST API

./ai-orchestrator run "Create a REST API with user authentication using JWT tokens"

Expected Flow:

  1. Codex implements the initial REST API with authentication
  2. Gemini reviews for security best practices and SOLID principles
  3. Claude implements the feedback and improvements

Example 2: Implement a Data Structure

./ai-orchestrator run "Implement a binary search tree with insert, delete, and search operations" --workflow thorough

Expected Flow:

  1. Codex creates initial implementation
  2. Copilot suggests optimizations
  3. Gemini reviews data structure design
  4. Claude implements improvements
  5. Gemini re-reviews

Example 3: Quick Implementation

./ai-orchestrator run "Create a function to validate email addresses" --workflow quick

Expected Flow:

  1. Codex implements quickly (no review cycle)

Advanced Usage

Specifying Maximum Iterations

./ai-orchestrator run "Refactor the authentication module" --max-iterations 5

Using Specific Output Directory

./ai-orchestrator run "Create a CLI tool for file processing" --output ./my-project

Verbose Mode for Debugging

./ai-orchestrator run "Fix the memory leak in the caching layer" --verbose

Dry Run (See Execution Plan)

./ai-orchestrator run "Optimize database queries" --dry-run

Configuration Examples

Custom Workflow

Edit config/agents.yaml:

workflows:
  security-focused:
    - agent: "codex"
      task: "implement"
      description: "Initial implementation"

    - agent: "gemini"
      task: "review"
      description: "Security audit"

    - agent: "gemini"
      task: "review"
      description: "Second security pass"

    - agent: "claude"
      task: "refine"
      description: "Implement security improvements"

Use it:

./ai-orchestrator run "Create payment processing system" --workflow security-focused

Disable Specific Agent

Edit config/agents.yaml:

agents:
  copilot:
    enabled: false  # Temporarily disable Copilot

Sample Task Categories

1. New Feature Implementation

./ai-orchestrator run "Add a caching layer using Redis with TTL support"
./ai-orchestrator run "Implement OAuth2 authentication with Google and GitHub"
./ai-orchestrator run "Create a background job queue using Celery"

2. Code Refactoring

./ai-orchestrator run "Refactor the user service to follow SOLID principles" --workflow review-only
./ai-orchestrator run "Extract common functionality into reusable utilities"
./ai-orchestrator run "Improve error handling across the application"

3. Bug Fixes

./ai-orchestrator run "Fix the race condition in the order processing system"
./ai-orchestrator run "Resolve memory leak in the connection pool"
./ai-orchestrator run "Fix SQL injection vulnerability in search endpoint"

4. Testing

./ai-orchestrator run "Write comprehensive unit tests for the authentication module"
./ai-orchestrator run "Add integration tests for the payment processing workflow"
./ai-orchestrator run "Create end-to-end tests for user registration flow"

5. Documentation

./ai-orchestrator run "Generate API documentation with examples" --workflow document
./ai-orchestrator run "Create user guide for the CLI tool"
./ai-orchestrator run "Document the deployment process with diagrams"

CLI Commands Reference

Run a Task

./ai-orchestrator run <task> [options]

Options:

  • --workflow, -w: Workflow to use (default, quick, thorough, etc.)
  • --config, -c: Path to configuration file
  • --max-iterations, -m: Maximum refinement iterations
  • --output, -o: Output directory
  • --verbose, -v: Enable verbose logging
  • --dry-run: Show execution plan without running

List Available Agents

./ai-orchestrator agents

List Available Workflows

./ai-orchestrator workflows

Validate Configuration

./ai-orchestrator validate

Show Version

./ai-orchestrator version

Expected Output Examples

Successful Execution

╭─────────────────────────────────────────────╮
│ AI Orchestrator                             │
│ Task: Create a REST API with authentication │
│ Workflow: default                           │
╰─────────────────────────────────────────────╯

Available agents: codex, gemini, claude

Starting execution...

Iteration 1:
  ✓ codex - implement
     Files modified: 3
  ✓ gemini - review
     Suggestions: 5
  ✓ claude - refine
     Files modified: 3

Iteration 2:
  ✓ codex - implement
  ✓ gemini - review
     Suggestions: 1
  ✓ claude - refine

Execution complete! ✓

Task completed successfully! ✓

Agent Availability Check

╭─────────────────────────────────────────────╮
│ AI Agents                                   │
├─────────┬──────────────┬────────────┬───────╯
│ Agent   │ Status       │ Command    │ Role
├─────────┼──────────────┼────────────┼─────────
│ codex   │ ✓ Available  │ codex      │ implementation
│ gemini  │ ✓ Available  │ gemini-cli │ review
│ claude  │ ✓ Available  │ claude     │ refinement
│ copilot │ ○ Disabled   │ gh         │ suggestions
╰─────────┴──────────────┴────────────┴─────────

Tips and Best Practices

1. Start with Dry Run

Before running a complex task, use --dry-run to see the execution plan:

./ai-orchestrator run "Complex task" --dry-run

2. Use Appropriate Workflows

  • quick: Simple tasks, no review needed
  • default: Standard implementation-review-refine cycle
  • thorough: Multiple review passes for critical code
  • review-only: For existing code that needs improvement

3. Monitor Logs

Check ai-orchestrator.log for detailed execution information:

tail -f ai-orchestrator.log

4. Validate Configuration

After modifying configuration, validate it:

./ai-orchestrator validate

5. Iterative Refinement

Use --max-iterations to control refinement depth:

  • 1-2 iterations: Quick tasks
  • 3-4 iterations: Standard tasks
  • 5+ iterations: Complex, critical code

Troubleshooting

No Agents Available

Problem: "No agents available! Please install and configure AI CLI tools."

Solution:

  1. Install required CLI tools
  2. Authenticate with each service
  3. Verify with ./ai-orchestrator agents

Timeout Errors

Problem: Tasks timing out

Solution:

  1. Increase timeout in config/agents.yaml
  2. Break task into smaller sub-tasks
  3. Use --max-iterations to limit refinement

Configuration Errors

Problem: Invalid configuration

Solution:

./ai-orchestrator validate

Fix issues reported and retry.

Real-World Examples

Example 1: Building a Complete Feature

# Step 1: Implement core functionality
./ai-orchestrator run "Create a task queue system with priority support"

# Step 2: Add tests
./ai-orchestrator run "Write unit and integration tests for task queue"

# Step 3: Add documentation
./ai-orchestrator run "Generate documentation for task queue API" --workflow document

Example 2: Code Quality Improvement

# Review existing code
./ai-orchestrator run "Review and refactor src/services/payment.py" --workflow review-only

# Implement improvements
./ai-orchestrator run "Add error handling and input validation to payment service"

# Final security review
./ai-orchestrator run "Security audit of payment processing" --workflow thorough

Example 3: Testing and Quality Assurance

# Generate tests
./ai-orchestrator run "Create comprehensive test suite for user authentication"

# Review test coverage
./ai-orchestrator run "Review and improve test coverage for authentication module" --workflow review-only