agcluster-container

🥈Silver

Self-hosted platform for Claude Agent SDK agents. Container isolation, real-time execution tracking, and Web UI dashboard. Operations teams can deploy and manage AI agents in isolated environments, monitor performance, and integrate with existing workflows.

70100Updated 1mo ago

Intermediate30min to implementautomation

Saves ~150 min per use

Quick InstallView Source

git clone https://github.com/whiteboardmonk/agcluster-container.git

Works with:

Claude

Overview

About This Skill

How to Use

[{"step":"Install agcluster-container on a dedicated server or VM using the official installation guide for your OS (e.g., `curl -sSL https://get.agcluster.io | bash`).","tip":"Use a cloud instance with at least 8GB RAM and 4 vCPUs for testing."},{"step":"Define agent configurations in a YAML file (e.g., `agents.yml`) specifying container images, resource limits, and environment variables.","tip":"Start with minimal resources and scale up based on agent performance. Use pre-built images from `agcluster/agent-templates` for common tasks."},{"step":"Deploy agents using the CLI: `agcluster-cli deploy --config agents.yml`. Verify deployment status with `agcluster-cli status`.","tip":"Use `agcluster-cli logs --agent <agent-name>` to debug startup issues."},{"step":"Configure real-time tracking in the Web UI by navigating to `http://<server-ip>:8080/dashboard` and setting up metrics for your agents.","tip":"Enable Prometheus integration if you use Grafana for dashboards. Pre-configured dashboards are available in the `agcluster/dashboards` repo."},{"step":"Integrate with existing workflows by adding deployment steps to your CI/CD pipeline (e.g., GitHub Actions) or configuring webhooks for agent triggers.","tip":"Use the `agcluster-cli` API for programmatic control (e.g., `agcluster-cli restart --agent <agent-name>`)."}]

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/whiteboardmonk/agcluster-container

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

Deploy and configure the agcluster-container platform to host [NUMBER] AI agents for [TEAM/PROJECT]. Ensure each agent runs in an isolated container with [RESOURCE_LIMITS] (e.g., CPU/memory). Set up real-time execution tracking for [AGENTS_TO_MONITOR] and configure the Web UI dashboard to display [METRICS_TO_TRACK] (e.g., latency, error rates). Provide step-by-step instructions for integration with [EXISTING_TOOLS] (e.g., CI/CD pipelines, monitoring tools).

Example Output

### agcluster-container Deployment Report
**Platform Version:** agcluster-container v2.1.0
**Deployment Date:** 2024-05-20
**Agents Deployed:** 5 (isolated containers)

#### Container Configuration:
- **Agent 1 (DataProcessor):** 2 vCPUs, 4GB RAM, Ubuntu 22.04
- **Agent 2 (NLP-Analyzer):** 4 vCPUs, 8GB RAM, Debian 11
- **Agent 3 (ImageClassifier):** 2 vCPUs, 6GB RAM, Alpine Linux
- **Agent 4 (SentimentBot):** 1 vCPU, 2GB RAM, Ubuntu 22.04
- **Agent 5 (WorkflowOrchestrator):** 3 vCPUs, 10GB RAM, Debian 11

#### Real-Time Monitoring Setup:
- **Execution Tracking:** Enabled for all agents via Prometheus metrics endpoint (`http://agcluster.local:9090/metrics`).
- **Dashboard Metrics:** Latency (avg. 120ms), Error Rate (<0.5%), CPU/Memory Usage (plotted hourly).
- **Alerts:** Configured for high latency (>500ms) or memory spikes (>90%).

#### Integration with Existing Tools:
1. **CI/CD Pipeline (GitHub Actions):** Added a step to deploy agents post-build using `agcluster-cli deploy --config agents.yml`.
2. **Monitoring (Grafana):** Connected to Prometheus for visualization. Created a dashboard with agent-specific panels.
3. **Logging (ELK Stack):** Forwarded container logs to Elasticsearch for analysis.

#### Next Steps:
- Test agent interactions via the Web UI at `http://agcluster.local/dashboard`.
- Scale resources for Agent 5 (WorkflowOrchestrator) if CPU usage exceeds 70% for 1 hour.
- Schedule a review of error logs weekly to identify recurring issues.

**Status:** All agents operational. No critical alerts triggered in the last 24 hours.