| layout | single | ||
|---|---|---|---|
| title | API Reference | ||
| description | Complete CLI and Python API documentation | ||
| toc | true | ||
| toc_label | API Topics | ||
| toc_icon | code | ||
| sidebar |
|
AI Command Auditor provides both a powerful command-line interface (CLI) and a comprehensive Python API for integration into your workflows and applications.
The CLI is the primary way to interact with AI Command Auditor. It provides commands for validation, configuration, and integration.
# Core Commands
ai-auditor check-command "your command here" # Validate a command
ai-auditor init # Initialize in project
ai-auditor setup-hooks # Setup git hooks
# Configuration
ai-auditor config show # View configuration
ai-auditor config set security.strict_mode true # Update settings
# Information
ai-auditor --version # Show version
ai-auditor --help # Show helpThe Python API allows you to integrate AI Command Auditor into your Python applications, scripts, and automation workflows.
from ai_command_auditor import CommandAuditor, SecurityRules
# Initialize auditor
auditor = CommandAuditor()
# Check a command
result = auditor.analyze_command("rm -rf /tmp/*")
print(f"Safety Score: {result.safety_score}")
print(f"Recommendations: {result.recommendations}")
# Load custom rules
rules = SecurityRules.from_file("custom-rules.yml")
auditor.update_rules(rules)# 1. Install and initialize
curl -fsSL https://raw.githubusercontent.com/etherisc/ai-command-auditor/main/install.sh | sh
cd your-project && ai-auditor init
# 2. Check your first command
ai-auditor check-command "sudo rm -rf /var/log/*"
# 3. Setup git hooks
ai-auditor setup-hooks
# 4. Configure for your project
ai-auditor config set templates.default "python"# 1. Install the package
# pip install ai-command-auditor
# 2. Basic usage
from ai_command_auditor import CommandAuditor
auditor = CommandAuditor()
result = auditor.analyze_command("python -c 'import os; os.system(\"rm -rf /\")'")
if result.is_dangerous:
print(f"⚠️ Dangerous command detected!")
print(f"Severity: {result.severity}")
print(f"Reason: {result.reason}")
else:
print("✅ Command appears safe")
# 3. Custom configuration
auditor.configure({
'security': {'strict_mode': True},
'ai': {'model': 'gpt-4o'}
})The complete CLI reference covers all commands, options, and use cases:
| Section | Description | Link |
|---|---|---|
| Command Validation | Validate commands and scripts | CLI Commands → |
| Configuration Management | Manage settings and templates | Config Commands → |
| Git Integration | Setup hooks and automation | Git Commands → |
| Reporting & Analysis | Generate reports and insights | Report Commands → |
| Utility Commands | Helper and diagnostic commands | Utility Commands → |
Complete Python API reference with examples:
| Module | Description | Link |
|---|---|---|
CommandAuditor |
Main auditor class and methods | Core API → |
SecurityRules |
Security policy management | Security API → |
AIAnalyzer |
AI-powered analysis engine | AI API → |
Configuration |
Configuration management | Config API → |
Validators |
Custom validation framework | Validators → |
# GitHub Actions example
name: Command Validation
on: [push, pull_request]
jobs:
validate:
runs-on: ubuntu-latest
steps:
- uses: actions/checkout@v3
- name: Install AI Command Auditor
run: curl -fsSL https://raw.githubusercontent.com/etherisc/ai-command-auditor/main/install.sh | sh
- name: Validate Commands
run: |
ai-auditor scan-scripts scripts/
ai-auditor validate-makefile Makefile# .pre-commit-config.yaml
repos:
- repo: local
hooks:
- id: ai-command-auditor
name: AI Command Auditor
entry: ai-auditor check-staged-commands
language: system
stages: [commit]# Custom integration example
import subprocess
from ai_command_auditor import CommandAuditor
class SafeCommandRunner:
def __init__(self):
self.auditor = CommandAuditor()
def run_command(self, command: str) -> str:
# Validate before execution
result = self.auditor.analyze_command(command)
if result.is_dangerous:
raise SecurityError(f"Dangerous command blocked: {result.reason}")
# Execute if safe
return subprocess.check_output(command, shell=True, text=True)
# Usage
runner = SafeCommandRunner()
output = runner.run_command("ls -la") # Safe
# runner.run_command("rm -rf /") # Would raise SecurityErrorfrom ai_command_auditor.validators import BaseValidator
class CustomValidator(BaseValidator):
def __init__(self):
super().__init__("custom_validator")
def validate(self, command: str) -> ValidationResult:
# Custom validation logic
if "dangerous_pattern" in command:
return ValidationResult(
valid=False,
severity="high",
message="Custom security rule violation"
)
return ValidationResult(valid=True)
# Register custom validator
auditor = CommandAuditor()
auditor.add_validator(CustomValidator())# Create a plugin
from ai_command_auditor.plugins import BasePlugin
class MyPlugin(BasePlugin):
def on_command_analyzed(self, command: str, result: AnalysisResult):
# Custom post-analysis logic
if result.severity == "critical":
self.send_alert(command, result)
def send_alert(self, command: str, result: AnalysisResult):
# Send notification, log to external system, etc.
pass
# Load plugin
auditor.load_plugin(MyPlugin())import asyncio
from ai_command_auditor import AsyncCommandAuditor
async def validate_commands(commands: list[str]):
auditor = AsyncCommandAuditor()
# Validate multiple commands concurrently
results = await asyncio.gather(*[
auditor.analyze_command(cmd) for cmd in commands
])
return results
# Usage
commands = ["ls -la", "rm temp.txt", "python script.py"]
results = asyncio.run(validate_commands(commands))from ai_command_auditor import CommandAuditor
# Configure caching for better performance
auditor = CommandAuditor({
'performance': {
'cache_enabled': True,
'cache_ttl': 3600, # 1 hour
'parallel_requests': 5
}
})
# Batch processing
results = auditor.analyze_commands(command_list, batch_size=10)# Use custom AI model
auditor = CommandAuditor({
'ai': {
'model': 'custom-model',
'api_base': 'https://your-api.com/v1',
'api_key': 'your-api-key'
}
})from ai_command_auditor import SecurityPolicy
# Create custom security policy
policy = SecurityPolicy()
policy.add_rule("no_rm_rf", r"rm\s+-rf\s+/", severity="critical")
policy.add_rule("no_curl_eval", r"curl.*\|\s*sh", severity="high")
auditor = CommandAuditor()
auditor.apply_policy(policy)- 🖥️ Complete CLI Reference - All CLI commands and options
- 🐍 Complete Python API - Full Python API documentation
- 🔗 Integration Guide - Integration patterns and examples
- 👩💻 Developer Guide - Contributing and development
- Installation Guide - Get started quickly
- Configuration Reference - Configure for your needs
- Examples & Tutorials - Learn with practical examples
- Support & FAQ - Get help and find answers