AgentOS for Copilot
Build safe AI agents with natural language and 0% policy violations
Create agents from descriptions, use 50+ templates, ensure compliance with GDPR, HIPAA, SOC2, and deploy directly to GitHub Actionsβall through Copilot Chat.
β‘ Quick Commands
Use these commands in GitHub Copilot Chat
@agentos create agent for [task]
Generate agent from natural language
@agentos templates
Browse 50+ pre-built templates
@agentos compliance gdpr
Check GDPR compliance
@agentos deploy
Deploy to GitHub Actions
π― Features & Tutorials
Click any feature to learn how to use it
Getting Started
Install the extension, run your first command, and create a working agent in under 5 minutes with step-by-step guidance.
Agent Creation
Create agents from natural language descriptions. Automatic task decomposition, policy detection, and code generation in Python, TypeScript, or Go.
Template Gallery
50+ pre-built templates across 8 categories: Data Processing, DevOps, Customer Support, Content, BI, Security, Integration, Automation.
Compliance Frameworks
Built-in GDPR, HIPAA, SOC 2, and PCI DSS compliance checking. Automatic policy generation and violation detection.
Testing & Security
Automated test generation, edge case detection, security vulnerability scanning, and cost estimation before deployment.
Debugging & Optimization
AI-powered error diagnosis, execution trace analysis, performance bottleneck detection, and automated fix suggestions.
GitHub Deployment
One-command deployment to GitHub Actions. Auto-generates workflows, manages secrets, creates PRs with agent code and policies.
Multi-Model Review
CMVK (Cross-Model Verification Kernel) reviews code with multiple AI models for consensus-based safety verification.
π Quick Preview
User: @agentos create agent for processing customer feedback from Slack
AgentOS: π€ Agent Created: CustomerFeedbackProcessingAgent
Processing customer feedback from Slack
### Tasks
- Read data from Slack API
- Analyze and process data
- Send results to Database
### π‘οΈ Safety Policies Applied
β
API Rate Limiting (rate_limit): Limits API calls to prevent quota exhaustion
β
PII Protection (data_privacy): Redacts personally identifiable information
β
Authentication Required (auth): Requires valid API token
β
Audit Logging (logging): Logs all agent actions for audit trail
### Generated Code (Python)
```python
from agent_os import KernelSpace
from agent_os.integrations import create_safe_toolkit
kernel = KernelSpace(policy="strict")
toolkit = create_safe_toolkit("standard")
SLACK_TOKEN = os.environ.get("SLACK_TOKEN")
@kernel.register
async def customer_feedback_processing_agent(task: str):
results = {"status": "success", "timestamp": datetime.utcnow().isoformat()}
data = await fetch_from_slack_api()
processed = await process_data(data)
await send_to_database(processed)
return results
```
[π Deploy to GitHub Actions] [π§ͺ Test Agent] [π Copy Code]
π Complete Command Reference
| Command | Description | Example |
|---|---|---|
@agentos create |
Create agent from description | @agentos create agent for monitoring API uptime |
@agentos design |
Design multi-step workflow | @agentos design workflow to generate reports |
@agentos templates |
Browse template gallery | @agentos templates devops |
@agentos test |
Test agent with scenarios | @agentos test |
@agentos debug |
Debug agent failures | @agentos why did this agent fail? |
@agentos compliance |
Check compliance framework | @agentos compliance hipaa |
@agentos security |
Run security audit | @agentos security |
@agentos deploy |
Deploy to GitHub Actions | @agentos deploy |
@agentos optimize |
Performance optimization | @agentos optimize |
@agentos review |
CMVK multi-model review | @agentos review |
@agentos policy |
Show active policies | @agentos policy |
@agentos audit |
View audit log | @agentos audit |
@agentos help |
Show all commands | @agentos help |
Ready to get started?
π Getting Started Tutorial β