kuberocketai

🥈Silver

KubeRocketAI is a declarative framework for AI-driven software development. It enables teams to define, validate, and orchestrate AI agents as code. This makes the software development lifecycle (SDLC) transparent, auditable, and CI/CD-ready. It integrates with tools like VSCode, Cursor, and Windsurf, benefiting operations teams by automating and streamlining software development workflows.

3160Updated 1mo ago

Intermediate30min to implementautomation

Saves ~140 min per use

Quick InstallView Source

git clone https://github.com/KubeRocketCI/kuberocketai.git

Works with:

Claude

Overview

About This Skill

How to Use

[{"step":"Define the AI agent's purpose and requirements","action":"Use the prompt template to generate a manifest. Replace [TASK_DESCRIPTION] with the agent's role (e.g., 'automate code reviews'), [MODEL_NAME] with the model to use (e.g., 'mistral-7b-instruct'), and [ADDITIONAL_RESOURCES] with any dependencies (e.g., 'PostgreSQL database for storing review history').","tip":"Use `kuberocketai list-models` to see available models in your cluster."},{"step":"Validate the manifest","action":"Run `kuberocketai validate -f code-review-agent.yaml` to check for syntax errors, resource conflicts, or permission issues. Fix any warnings before proceeding.","tip":"Enable verbose mode with `-v debug` to get detailed validation logs."},{"step":"Deploy the agent","action":"Apply the manifest to your cluster with `kuberocketai apply -f code-review-agent.yaml -n dev-team`. Monitor the deployment status with `kubectl get pods -n dev-team`.","tip":"Use `kuberocketai status -n dev-team` to get a high-level overview of all deployed agents."},{"step":"Integrate with development tools","action":"Configure the agent to interact with your SDLC tools (e.g., GitHub, GitLab, Jira). Use the agent's output configuration to specify where results should be posted (e.g., PR comments, Slack channels).","tip":"For GitHub, use the `github-webhook` integration in the manifest to trigger the agent on PR events."},{"step":"Monitor and iterate","action":"Use the monitoring dashboard (e.g., Grafana) to track the agent's performance. Adjust parameters like `temperature` or `max_tokens` in the manifest to improve output quality.","tip":"Set up alerts in the manifest to notify your team of issues (e.g., high latency or errors)."}]

Setup & Installation

Quick Install

No install command available. Check the GitHub repository for manual installation instructions.

Alternative Install (Git Clone)

git clone https://github.com/KubeRocketCI/kuberocketai

Requirements

Claude Code or compatible AI agent
Works with: Claude

Quick Start Guide

Install the Skill

Copy the install command above and run it in your terminal.

Open Your AI Agent

Launch Claude Code, Cursor, or your preferred AI coding agent.

Try It Out

Use the prompt template or examples below to test the skill.

Customize

Adapt the skill to your specific use case and workflow.

Usage Examples

Prompt Template

Generate a KubeRocketAI manifest for deploying an AI agent that [TASK_DESCRIPTION]. The agent should use [MODEL_NAME] and include [ADDITIONAL_RESOURCES] like databases or APIs. Validate the manifest with `kuberocketai validate` and deploy it to a [NAMESPACE] using `kuberocketai apply`. Ensure the agent has [PERMISSIONS] and [MONITORING] configured.

Example Output

```yaml
apiVersion: kuberocket.ai/v1alpha1
kind: AIAgent
metadata:
  name: code-review-agent
  namespace: dev-team
spec:
  model:
    name: "mistral-7b-instruct"
    provider: "huggingface"
    parameters:
      temperature: 0.3
      max_tokens: 2048
  tasks:
    - name: "code-review"
      description: "Review pull requests for security vulnerabilities, style issues, and best practices."
      input:
        source: "github"
        repository: "acme-corp/web-app"
        branch: "main"
        file_patterns: ["*.py", "*.js"]
      output:
        format: "markdown"
        destination: "github-pr-comments"
  resources:
    memory: "4Gi"
    cpu: "2"
    gpu: "1"
  permissions:
    - apiGroups: [""]
      resources: ["pods", "services"]
      verbs: ["get", "list", "watch"]
    - apiGroups: ["batch"]
      resources: ["jobs"]
      verbs: ["create", "get"]
  monitoring:
    metrics:
      - name: "agent_latency"
        description: "Time taken to process a code review request"
      - name: "agent_errors"
        description: "Number of errors encountered during processing"
    alerts:
      - name: "high-latency-alert"
        threshold: 30000
        severity: "warning"
```

### Validation Output:
```
✅ Manifest validated successfully for 'code-review-agent'
⚠️ Warning: GPU resource request exceeds cluster capacity (available: 0)
🔧 Suggestion: Adjust `gpu: "1"` to `gpu: "0"` or request a GPU-enabled node pool.
```

### Deployment Output:
```
🚀 Deploying 'code-review-agent' to namespace 'dev-team'...
✅ Agent deployed successfully!
📊 Monitoring dashboard: http://grafana.dev-team.svc.cluster.local/d/kuberocketai/code-review-agent
🔗 GitHub integration: https://github.com/acme-corp/web-app/settings/integrations/kuberocketai
```