aitbc/docs/agent-training/ENVIRONMENT_SETUP.md

Training Environment Setup Guide

This guide explains how to set up the AITBC training environment using the Python-based setup system, which replaces the previous shell script approach.

Overview

The Python-based training environment setup provides:

• Testability: Can be unit tested and integrated tested
• Maintainability: Python code is easier to debug and maintain
• Integration: Uses existing AITBC patterns and utilities
• Type Safety: Better error handling with exceptions
• IDE Support: Autocomplete and type hints
• CI/CD Integration: Easier to integrate with pytest-based CI
• Schema-Driven Execution: JSON-based stage definitions for reproducible training
• Deterministic Wallet Naming: Predictable wallet names for consistent testing
• Transaction Hash Validation: Automatic extraction and validation of transaction hashes

Installation

The training setup module is located in aitbc/training_setup/. To use it, you need to either install the AITBC package or add the directory to your Python path.

Option 1: Install the package (recommended):

cd /opt/aitbc
pip install -e .

Option 2: Add to Python path:

export PYTHONPATH="/opt/aitbc:$PYTHONPATH"

Option 3: Use directly with path:

cd /opt/aitbc
python3 -c "import sys; sys.path.insert(0, '.'); from aitbc.training_setup import TrainingEnvironment"

Quick Start

Using the CLI

# Setup complete training environment
python -m aitbc.training_setup.cli setup

# Check prerequisites
python -m aitbc.training_setup.cli check

# Verify environment
python -m aitbc.training_setup.cli verify

# Fund a specific wallet
python -m aitbc.training_setup.cli fund-wallet my-wallet --password my-password

# Run a training stage from JSON schema
python -m aitbc.training_setup.cli run-stage /path/to/stage.json

Using Python API

from aitbc.training_setup import TrainingEnvironment

# Create environment
env = TrainingEnvironment(
    aitbc_dir="/opt/aitbc",
    log_dir="/var/log/aitbc/training-setup",
    faucet_amount=1000,
    genesis_allocation=10000,
)

# Setup full environment
results = env.setup_full_environment()
print(results)

# Fund a specific wallet
env.fund_training_wallet("training-wallet", "password123")

# Configure messaging
env.configure_messaging_auth("training-wallet", "password123")

# Verify environment
verification = env.verify_environment()
print(verification)

# Run a stage from JSON schema
result = env.run_stage_from_json("/path/to/stage.json")
print(result)

# Get deterministic wallet name
wallet_name = env.get_wallet_name(1)  # training-w1
print(f"Wallet name: {wallet_name}")

Using Pytest Fixtures

For testing and CI/CD integration, use the pytest fixtures:

import pytest
from aitbc.training_setup import TrainingEnvironment

def test_training_setup(training_env_mock):
    """Test with mocked environment"""
    result = training_env_mock.fund_training_wallet("test-wallet")
    assert result["status"] == "completed"

@pytest.mark.integration
def test_real_setup(training_env):
    """Test with real environment"""
    result = training_env.check_prerequisites()
    assert result is True

Schema-Driven Stage Execution

The training setup supports schema-driven stage execution using JSON definitions. This approach provides:

• Reproducibility: Each stage runs the same way with the same commands
• Validation: Expected conditions are automatically validated
• Transaction Tracking: Transaction hashes are extracted and logged
• Test Integration: Stages can be run as pytest tests

Stage JSON Schema

Note: The training stages use the official schema defined in training_schema.json. The schema uses an operations array with operation/parameters/expected_result/success_criteria format, not the older commands format.

Example from stage1_foundation.json:

{
  "stage": "stage1_foundation",
  "agent_type": "general",
  "training_data": {
    "operations": [
      {
        "operation": "wallet_create",
        "parameters": {
          "name": "training-wallet",
          "password": "training123"
        },
        "expected_result": {
          "status": "success",
          "wallet_id": "training-wallet"
        },
        "success_criteria": {
          "status": "success",
          "response_fields": ["wallet_id", "address"]
        }
      }
    ]
  },
  "validation": {
    "exam_tests": [
      {
        "test_name": "Create wallet",
        "operation": "wallet_create",
        "test_case": {
          "name": "exam-wallet",
          "password": "exam123"
        },
        "expected_output": {
          "status": "success",
          "wallet_id": "exam-wallet"
        }
      }
    ],
    "passing_score": 80
  }
}

Running Stages

CLI:

python -m aitbc.training_setup.cli run-stage /path/to/stage.json

Python API:

from aitbc.training_setup import TrainingEnvironment

env = TrainingEnvironment()
result = env.run_stage_from_json("/path/to/stage.json")
print(result)

Deterministic Wallet Naming

The training environment uses deterministic wallet naming for consistency:

# Get wallet names by index
env = TrainingEnvironment(wallet_prefix="training-w")
wallet1 = env.get_wallet_name(1)  # training-w1
wallet2 = env.get_wallet_name(2)  # training-w2
wallet3 = env.get_wallet_name(3)  # training-w3

This ensures predictable wallet names across different training sessions and environments.

Transaction Hash Validation

The stage runner automatically extracts transaction hashes from command output:

result = env.run_stage_from_json("/path/to/stage.json")
for cmd_result in result['commands']:
    if cmd_result.get('tx_hash'):
        print(f"Transaction: {cmd_result['tx_hash']}")

Components

TrainingEnvironment Class

The main class that manages training environment setup.

Methods:

• check_prerequisites() - Verify AITBC CLI and node are available
• create_genesis_allocation() - Create genesis wallet and initialize blockchain
• setup_faucet_wallet() - Setup and fund faucet wallet
• fund_training_wallet(wallet_name, password) - Fund a specific wallet
• generate_auth_token() - Generate authentication token
• configure_messaging_auth(wallet_name, password) - Configure messaging
• test_messaging_connectivity() - Test messaging service
• verify_environment() - Verify all components are configured
• setup_full_environment() - Run complete setup process

Exceptions

Custom exceptions for error handling:

• TrainingSetupError - Base exception for setup errors
• FundingError - Account funding failures
• MessagingError - Messaging configuration failures
• FaucetError - Faucet setup failures
• PrerequisitesError - Prerequisites not met

Pytest Fixtures

Available in tests/conftest.py:

• training_env - Session-scoped real environment fixture
• training_env_mock - Function-scoped mocked environment fixture
• mock_faucet_response - Mock faucet API response
• training_stage_data - Sample training stage data

Configuration

Environment Variables

The setup system respects standard AITBC environment variables:

• AITBC_DIR - AITBC installation directory (default: /opt/aitbc)
• DATA_DIR - Data directory
• LOG_DIR - Log directory

Custom Parameters

env = TrainingEnvironment(
    aitbc_dir="/custom/path",           # AITBC directory
    log_dir="/custom/logs",              # Log directory
    faucet_amount=1000,                 # Tokens per faucet request
    genesis_allocation=10000,           # Genesis allocation amount
)

Logging

Logs are written to /var/log/aitbc/training-setup/training_setup.log by default.

Log levels:

• INFO - Normal operations
• WARNING - Non-critical issues
• ERROR - Failures that prevent setup completion

Migration from Shell Scripts

The shell scripts in /opt/aitbc/scripts/training/ are maintained for backward compatibility but are deprecated.

Migration Guide:

Shell Script	Python Equivalent
setup_training_env.sh	TrainingEnvironment().setup_full_environment()
fund_accounts.sh	TrainingEnvironment().fund_training_wallet()
configure_messaging.sh	TrainingEnvironment().configure_messaging_auth()
setup_faucet.sh	Manual setup or use existing faucet service

Testing

Unit Tests

pytest tests/training/test_training_setup.py -v

Integration Tests

pytest tests/training/test_training_setup.py -v -m integration

With Mocked Environment

pytest tests/training/test_training_setup.py -v -k "mock"

Troubleshooting

Prerequisites Not Met

Error: TrainingSetupError: AITBC CLI not found

Solution: Ensure AITBC is installed at the specified directory:

ls /opt/aitbc/aitbc-cli

Funding Failures

Error: FundingError: Failed to fund wallet

Solution: Check that:

• Genesis wallet exists and is funded (pre-funded with 999,999,890 AIT)
• AITBC node is running
• Network connectivity is available

Note: Locally created wallets aren't automatically funded on-chain. The setup system uses the pre-funded genesis wallet as the funding source.

Messaging Failures

Error: Messaging configuration warnings

Solution: Messaging configuration is optional. If it fails, the setup continues with a warning. Core blockchain operations don't require messaging. Check logs in /var/log/aitbc/training-setup/training_setup.log for details.

Genesis Initialization Errors

Error: CLI errors when using --force flag for genesis initialization

Solution: Genesis block already exists, so initialization is automatically skipped. The setup system checks genesis wallet status instead of attempting initialization.

Best Practices

1. Use Mocked Environment for Tests: Use training_env_mock fixture for unit tests
2. Check Prerequisites First: Always call check_prerequisites() before setup
3. Verify After Setup: Call verify_environment() to confirm setup success
4. Handle Exceptions: Use try/except blocks with specific exception types
5. Use Session Fixture: Use training_env for expensive setup operations
6. Log Review: Check logs in /var/log/aitbc/training-setup/ for debugging

Examples

Complete Setup Script

#!/usr/bin/env python3
from aitbc.training_setup import TrainingEnvironment, TrainingSetupError

def main():
    try:
        env = TrainingEnvironment()
        
        # Check prerequisites
        env.check_prerequisites()
        
        # Setup full environment
        results = env.setup_full_environment()
        
        # Verify setup
        verification = env.verify_environment()
        
        print("Setup completed successfully")
        print(f"Results: {results}")
        print(f"Verification: {verification}")
        
    except TrainingSetupError as e:
        print(f"Setup failed: {e}")
        exit(1)

if __name__ == "__main__":
    main()

Selective Setup

from aitbc.training_setup import TrainingEnvironment

env = TrainingEnvironment()

# Only setup funding
env.create_genesis_allocation()
env.setup_faucet_wallet()
env.fund_training_wallet("training-wallet")

# Only setup messaging
env.configure_messaging_auth("training-wallet")
env.test_messaging_connectivity()

Test Fixture Usage

import pytest

def test_wallet_funding(training_env_mock):
    """Test wallet funding with mocked environment"""
    result = training_env_mock.fund_training_wallet("test-wallet")
    assert result["status"] == "completed"
    assert result["amount"] == 1000

def test_messaging_setup(training_env_mock):
    """Test messaging configuration"""
    result = training_env_mock.configure_messaging_auth("test-wallet")
    assert result["status"] == "completed"
    assert "token_file" in result

Related Documentation

• Training Schema - JSON schema for training data
• Stage 1 Foundation - Stage 1 training configuration
• Operations Audit - Operations coverage analysis
• Test Suite README - Testing infrastructure

Support

For issues or questions:

1. Check logs in /var/log/aitbc/training-setup/
2. Run with verbose logging
3. Review test examples in tests/training/
4. Check AITBC documentation in /opt/aitbc/docs/