Quick to Deploy
Install with a single command and deploy your first AI agent in minutes. YAML-based configuration makes it easy to version control and share.
Flexible Tools
Integrate with kubectl, shell commands, HTTP APIs, Slack, GitHub, PagerDuty, and custom MCP tools. Build agents that work with your existing infrastructure.
Multi-Provider AI
Use OpenAI, Anthropic, Ollama, or Groq models. Switch providers easily and run locally or in the cloud.
Safe & Controlled
Human-in-the-loop approvals, allowed command lists, and audit logging keep your infrastructure safe while enabling automation.
Memory & Context
Persistent memory, RAG integration, and conversation history help agents learn and maintain context across interactions.
Production Ready
Built with Rust for performance and reliability. Supports fleets, workflows, scheduling, webhooks, and complex orchestration patterns.
Build Your First Agent in Minutes
# Install AOF
curl -sSL https://aof.dev/install.sh | bash
# Create an agent
cat > k8s-agent.yaml <<EOF
apiVersion: aof.dev/v1
kind: Agent
metadata:
name: k8s-helper
spec:
model: openai:gpt-4
instructions: |
You are a Kubernetes expert assistant.
Help users manage their clusters.
tools:
- type: Shell
config:
allowed_commands: [kubectl, helm]
EOF
# Run it
aofctl agent run k8s-agent.yamlBuilt for DevOps, SRE, and Platform Engineering
Incident Response
Automate diagnostics, remediation, and post-incident analysis. Integrate with PagerDuty, Slack, and your monitoring stack.
Kubernetes Management
Build intelligent K8s assistants that understand your cluster, troubleshoot issues, and execute safe operations.
Workflow Automation
Create multi-step workflows with AI agents that handle scheduling, webhooks, and complex decision-making.