Metadata-Version: 2.4
Name: vaikora
Version: 0.2.0
Summary: Python SDK for Vaikora AI Agent Security Platform
Project-URL: Homepage, https://vaikora.com
Project-URL: Documentation, https://docs.vaikora.com
Project-URL: Repository, https://github.com/zakrs/vaikora-python
Project-URL: Changelog, https://github.com/zakrs/vaikora-python/blob/main/CHANGELOG.md
Author-email: ZAKRS LLC <support@zakrs.com>
License-Expression: MIT
Keywords: agents,ai,autonomous,governance,llm,security
Classifier: Development Status :: 4 - Beta
Classifier: Intended Audience :: Developers
Classifier: License :: OSI Approved :: MIT License
Classifier: Programming Language :: Python :: 3
Classifier: Programming Language :: Python :: 3.9
Classifier: Programming Language :: Python :: 3.10
Classifier: Programming Language :: Python :: 3.11
Classifier: Programming Language :: Python :: 3.12
Classifier: Topic :: Security
Classifier: Topic :: Software Development :: Libraries :: Python Modules
Classifier: Typing :: Typed
Requires-Python: >=3.9
Requires-Dist: httpx>=0.25.0
Requires-Dist: pydantic>=2.0.0
Requires-Dist: python-dateutil>=2.8.0
Requires-Dist: typing-extensions>=4.0.0
Provides-Extra: dev
Requires-Dist: mypy>=1.0.0; extra == 'dev'
Requires-Dist: pytest-asyncio>=0.21.0; extra == 'dev'
Requires-Dist: pytest-cov>=4.0.0; extra == 'dev'
Requires-Dist: pytest>=7.0.0; extra == 'dev'
Requires-Dist: respx>=0.20.0; extra == 'dev'
Requires-Dist: ruff>=0.1.0; extra == 'dev'
Description-Content-Type: text/markdown

# Vaikora Python SDK

The official Python SDK for the Vaikora AI Agent Security Platform.

[![PyPI version](https://badge.fury.io/py/vaikora.svg)](https://badge.fury.io/py/vaikora)
[![Python Version](https://img.shields.io/pypi/pyversions/vaikora)](https://pypi.org/project/vaikora/)
[![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](https://opensource.org/licenses/MIT)

## Installation

```bash
pip install vaikora
```

## Quick Start

### Simple Configuration

```python
import vaikora

# Configure the SDK (do once at startup)
vaikora.configure(api_key="vk_your_api_key")

# Register your AI agent
agent = await vaikora.register(
    name="my-ai-agent",
    agent_type="autonomous",
    capabilities=["database_write", "api_call"]
)

print(f"Agent registered: {agent.id}")
```

### Using the Interceptor Decorator

The SDK provides a powerful decorator for securing AI agent actions:

```python
import vaikora

vaikora.configure(api_key="vk_your_api_key")

# Register agent first
await vaikora.register(name="my-agent", agent_type="autonomous")

@vaikora.interceptor(
    action_type="database.write",
    resource="production",
    require_approval=True,
    validate_inputs=True,
)
async def write_to_database(data: dict) -> bool:
    # Your code here - Vaikora validates and checks policies first
    await database.insert(data)
    return True

# Usage - automatically secured
result = await write_to_database({"user": "john", "action": "update"})
```

### Data Validation

Validate data for PII, anomalies, and toxicity:

```python
import vaikora

vaikora.configure(api_key="vk_your_api_key")

# Validate data before processing
result = await vaikora.validate_data(
    data={"email": "user@example.com", "ssn": "123-45-6789"},
    check_pii=True,
    check_anomalies=True,
    check_toxicity=True,
)

if result.pii_detected:
    print(f"PII found: {[p.pii_type for p in result.pii_detections]}")
    # Use cleaned data
    clean_data = result.cleaned_data
```

### Full Client Usage

For more control, use the client directly:

```python
from vaikora import VaikoraClient

# Initialize the client
client = VaikoraClient(api_key="vk_your_api_key")

# Register an agent
agent = await client.agents.register(
    name="my-ai-agent",
    agent_type="autonomous",
    capabilities=["database_write", "api_call", "file_access"]
)

# Submit an action for policy evaluation
result = await client.actions.submit(
    agent_id=str(agent.id),
    action_type="database_write",
    resource="users_table",
    payload={"user_id": 123, "data": {"name": "John"}}
)

if result.approved:
    # Proceed with the action
    pass
else:
    print(f"Action blocked: {result.denial_reason}")
```

## Synchronous Usage

For synchronous code, use the sync client:

```python
from vaikora import VaikoraClientSync

client = VaikoraClientSync(api_key="your-api-key")

# All methods are synchronous
agent = client.agents.register(
    name="my-sync-agent",
    agent_type="autonomous"
)
```

## Features

- **Action Control**: Submit actions for policy evaluation before execution
- **Anomaly Detection**: Automatic detection of unusual agent behavior
- **Policy Enforcement**: Server-side policy evaluation with local caching
- **Alerting**: Real-time alerts for security events
- **Audit Logging**: Immutable action logs with hash chains

## Configuration

```python
from vaikora import VaikoraClient

client = VaikoraClient(
    api_key="your-api-key",
    base_url="https://api.vaikora.com",  # Default
    timeout=30.0,  # Request timeout in seconds
    retry_count=3,  # Number of retries for failed requests
    agent_id="default-agent-id",  # Default agent for actions
)
```

## API Reference

### Agents

```python
# Register a new agent
agent = await client.agents.register(
    name="my-agent",
    agent_type="autonomous",  # autonomous, semi_autonomous, or supervised
    capabilities=["database_write", "api_call"],
    metadata={"version": "1.0"}
)

# Get agent by ID
agent = await client.agents.get(agent_id)

# List all agents
agents = await client.agents.list(page=1, page_size=20)

# Update agent
agent = await client.agents.update(agent_id, name="new-name")

# Deactivate agent
await client.agents.deactivate(agent_id)
```

### Actions

```python
# Submit action for evaluation
result = await client.actions.submit(
    agent_id="agent-uuid",
    action_type="database_write",
    resource="users_table",
    payload={"query": "UPDATE users SET name='John' WHERE id=123"},
    metadata={"source": "user_request"}
)

# Check result
if result.approved:
    print(f"Action approved: {result.action_id}")
else:
    print(f"Action denied: {result.denial_reason}")

# Get action details
action = await client.actions.get(action_id)

# List actions with filters
actions = await client.actions.list(
    agent_id="agent-uuid",
    action_type="database_write",
    status="approved",
    start_date=datetime(2024, 1, 1),
    end_date=datetime.now()
)
```

### Policies

```python
# List policies
policies = await client.policies.list()

# Get policy details
policy = await client.policies.get(policy_id)
```

### Alerts

```python
# List alerts
alerts = await client.alerts.list(status="open", severity="high")

# Get alert details
alert = await client.alerts.get(alert_id)

# Acknowledge alert
await client.alerts.acknowledge(alert_id)

# Resolve alert
await client.alerts.resolve(alert_id, notes="Issue addressed")
```

## Error Handling

```python
from vaikora import VaikoraClient
from vaikora.exceptions import (
    VaikoraError,
    AuthenticationError,
    RateLimitError,
    PolicyViolationError,
    NetworkError,
)

client = VaikoraClient(api_key="your-api-key")

try:
    result = await client.actions.submit(...)
except AuthenticationError:
    print("Invalid API key")
except RateLimitError as e:
    print(f"Rate limited. Retry after {e.retry_after} seconds")
except PolicyViolationError as e:
    print(f"Policy violation: {e.policy_name} - {e.message}")
except NetworkError:
    print("Network error. Check your connection.")
except VaikoraError as e:
    print(f"Vaikora error: {e}")
```

## License

MIT License - see [LICENSE](LICENSE) for details.
