01Agents

A Kubernetes alert remediation system with a modern web interface and an intelligent remediation agent. The platform analyzes monitoring alerts, retrieves live cluster context, and generates executable remediation scripts through LLM-powered workflows. It leverages intelligent tools and strategies to gather relevant system information and incorporates historical context to improve diagnosis and decision-making. Additionally, it supports advanced monitoring and telemetry capabilities to provide deeper visibility into cluster health and performance.

📁 Repository Structure

• helm-chart/: Contains the core Helm chart for deploying the agents and their mandatory dependencies.
• README.md: Main setup guide for getting the agents up and running (this file).
• OTEL-setup.md: Optional guide for setting up monitoring (Prometheus, Grafana, Loki, Tempo).

---

🚀 Quick Start Guide

To get the agents running, you need three mandatory steps:

1. Install MCP Server for Kubernetes

The MCP (Model Context Protocol) Server provides Kubernetes tools for the agents.

Clone the repository

git clone https://github.com/Flux159/mcp-server-kubernetes.git

Run the python script to update the schema (Required before install)

python3 -c "
import json
with open('mcp-server-kubernetes/helm-chart/values.schema.json') as f:
    schema = json.load(f)

schema['properties']['observability'] = {
    'type': 'object',
    'additionalProperties': True
}

with open('mcp-server-kubernetes/helm-chart/values.schema.json', 'w') as f:
    json.dump(schema, f, indent=2)
print('Done')
"

Install MCP Server

helm install mcp-server ./mcp-server-kubernetes/helm-chart \
  --set kubeconfig.provider=serviceaccount \
  --set transport.mode=http \
  --set transport.service.type=ClusterIP \
  --set security.allowOnlyNonDestructive=false \
  --create-namespace \
  --namespace mcp-system

2. Install CrunchyData PostgreSQL Operator

The agents use PostgreSQL for long-term memory. We use the CrunchyData Operator to manage it.

Install OLM (Operator Lifecycle Manager)

curl -sL https://github.com/operator-framework/operator-lifecycle-manager/releases/download/v0.40.0/install.sh | bash -s v0.40.0

Install the PostgreSQL Operator

kubectl create -f https://operatorhub.io/install/postgresql.yaml

3. Deploy 01Cloud Agents

Install the agents using this Helm chart. This will also automatically create the PostgreSQL cluster.

helm upgrade --install 01cloud-agent ./helm-chart -n 01cloud --create-namespace

---

4. Update Agent Configuration

Modify values.yaml to configure your agents. You must set the MODEL_PROVIDER, MODEL_NAME, and provide the corresponding API key in the secret section.

agents:
  - name: l0
    enabled: true
    image: myregistry/l0-agent:latest
    env:
      MODEL_PROVIDER: deepseek # options: gemini, openai, openrouter, anthropic, deepseek
      MODEL_NAME: deepseek-chat # examples: gemini-2.0-flash, gpt-4o, claude-3-5-sonnet
      MCP_SERVER_URL: http://mcp-server-mcp-server-kubernetes.mcp-system.svc.cluster.local:3001/mcp
      ENABLE_K8S_TOOLS: "true"
      STM_ENABLE_POSTGRES: "true"
    usePostgresql: true
    secret:
      DEEPSEEK_API_KEY: "your-api-key"
      # GOOGLE_API_KEY: "your-api-key"
      # OPENAI_API_KEY: "your-api-key"
      # OPENROUTER_API_KEY: "your-api-key"
      # ANTHROPIC_API_KEY: "your-api-key"

Apply changes:

helm upgrade --install 01cloud-agent ./helm-chart -n 01cloud

5. Access the Application

Once the agents are deployed, you can access the agent UI by port-forwarding the agent-l0 service:

kubectl port-forward svc/agent-l0 -n 01cloud 3000:3000

Now you can open your browser and go to http://localhost:3000.

---

6. Monitoring & Observability (Optional)

For advanced monitoring (Grafana, Loki, Tempo, OpenTelemetry), see OTEL-setup.md.

Local Access (Tools)

The following services are available via port-forwarding:

# PostgreSQL
kubectl port-forward svc/agents-primary -n 01cloud 5432:5432

# MCP Server
kubectl port-forward svc/mcp-server-mcp-server-kubernetes -n mcp-system 3001:3001