AgentStack
MCP verified MIT Self-run

Devparana Incident Agent Mcps

mcp-gabriel-dantas98-devparana-incident-agent-mcps · by gabriel-dantas98

Projeto criado para Talk no DevPR 2025

No reviews yet
0 installs
2 views
0.0% view→install

Install

$ agentstack add mcp-gabriel-dantas98-devparana-incident-agent-mcps

✓ scanned · ✓ verified — works with Claude Code, Cursor, and more.

Security review

✓ Passed

No issues found. Passed automated security review. · v0.1.0 How review works →

  • Prompt-injection patterns
  • Secret / credential exfiltration
  • Dangerous shell & filesystem operations
  • Untrusted network calls
  • Known-malicious package signatures

What it can access

  • Network access No
  • Filesystem access No
  • Shell / process execution No
  • Environment & secrets Used
  • Dynamic code execution No

From automated source analysis of v0.1.0. “Used” means the capability is present in the source — more access means more to trust, not that it’s unsafe.

Are you the author of Devparana Incident Agent Mcps? Claim this listing to set pricing, connect Stripe payouts, and keep 70% of every sale.
Sign up to claim

About

�� VibeDebugger - Incident Assistant for Oncallers with MCPs

AI-powered incident response assistant built with LangGraph ReAct agents and Model Context Protocol (MCP) for DevPR 2025 conference.

📊 Presentation: https://docs.google.com/presentation/d/1caZhQFIXv2K4e1ljIR9I85Xv51vegV3-S64tanx4Q4M/edit?usp=sharing

🎯 Quick Start

# 1. Setup Kubernetes cluster and all services
make setup

# 2. Configure local DNS (requires sudo)
make setup-hosts

# 3. Copy environment variables
cp .env.example .env  # Edit with your tokens

# 4. Install Python dependencies
uv install

# 5. Launch Discord bot
python -m src.vibedebugger_discord.chatbot

📋 Prerequisites

  • Docker (v20.10+) - Container runtime
  • Kind (v0.20+) - Local Kubernetes clusters
  • kubectl (v1.28+) - Kubernetes CLI
  • k9s (optional) - Terminal UI for Kubernetes
  • Python (3.10+) - Agent runtime
  • uv - Fast Python package manager
  • Node.js (v20) via nvm - For Backstage only
  • Discord Bot Token - For chat interface

🛠️ Available Commands

Main Makefile

make agent PROMPT="Your question"  # Run agent from terminal

Cluster Management (cluster/Makefile)

| Command | Description | |---------|-------------| | make setup | Create Kind cluster with all services (ArgoCD, Prometheus, Grafana, Backstage) | | make teardown | Destroy cluster and cleanup | | make setup-hosts | Configure *.localdev.me domains in /etc/hosts | | make status | Check health of all components | | make validate | Run validation tests | | make port-forward | Expose all services locally | | make get-passwords | Retrieve admin credentials | | make deploy-sample | Deploy nginx sample app for testing | | make logs- | View logs (argocd/prometheus/grafana/backstage/discord-bot) |

📁 Project Structure

hello-langrafo/
├── src/                    # Agent source code
│   ├── agent/             # LangGraph ReAct agent core
│   │   ├── graph.py       # Main agent orchestration
│   │   └── constants.py   # System prompts and configurations
│   ├── tools/             # MCP tool implementations  
│   │   ├── k8s.py        # Kubernetes operations
│   │   ├── prometheus.py  # Metrics and alerts
│   │   ├── argocd.py     # GitOps management
│   │   └── backstage.py  # Service catalog
│   ├── mcps/              # MCP server declarations
│   └── vibedebugger_discord/  # Discord bot interface
│       └── chatbot.py     # Bot implementation
├── cluster/               # Kubernetes infrastructure
│   ├── resources/        # K8s manifests
│   │   ├── argocd/      # GitOps deployment configs
│   │   ├── backstage/   # Developer portal
│   │   ├── grafana/     # Dashboards and visualizations  
│   │   ├── prometheus/  # Monitoring and alerting
│   │   └── sample-apps/ # Test applications
│   ├── scripts/         # Automation scripts
│   └── Makefile        # Cluster management commands
├── scripts/              # Helper utilities
│   └── agent_cli.py     # Terminal interface for agent
├── assets/              # Documentation images
├── docker-compose.yml   # Langfuse observability stack
├── langgraph.json      # LangGraph configuration
├── pyproject.toml      # Python dependencies
└── uv.lock            # Locked dependencies

🚀 Service Access

After make setup, services are available at:

| Service | URL | Default Credentials | |---------|-----|---------------------| | ArgoCD | http://argocd.localdev.me | admin / (run make get-passwords) | | Backstage | http://backstage.localdev.me | No auth required | | Grafana | http://grafana.localdev.me | admin / admin | | Prometheus | http://prometheus.localdev.me | No auth required |

> Note: Run make setup-hosts once to configure local DNS

💬 Usage Examples

Via Discord Bot

  1. Configure .env with your Discord bot token
  2. Run python -m src.vibedebugger_discord.chatbot
  3. Send messages in Discord channel

Via Terminal

# Direct query
make agent PROMPT="Help me investigate nginx-sample in apps namespace"

# Or using the CLI script
python scripts/agent_cli.py "Check pod status in all namespaces"

🔧 Configuration

Create .env file with:

# Required
OPENAI_API_KEY=sk-...
DISCORD_BOT_TOKEN=...
DISCORD_CHANNEL_ID=...

# Optional (uses defaults)
PROMETHEUS_URL=http://localhost:9090
ARGOCD_SERVER=localhost:8080
BACKSTAGE_URL=http://localhost:7007

📸 Demonstrations

Click to view screenshots

K9s Cluster Overview

Discord Alert from Prometheus

VibeDebugger Bot Response

🏗️ Architecture

  • ReAct Agent: Reasoning + Acting loop with LangGraph
  • MCP Tools: Direct integrations with K8s, Prometheus, ArgoCD, Backstage
  • Local Kind Cluster: Full observability stack for testing
  • Discord Interface: Real-time incident response chat

📝 License

MIT - See [LICENSE](LICENSE) file for details

🤝 Contributing

This project was created for DevPR Conference 2025. Feel free to fork and adapt for your needs!


Built with ❤️ by Gabriel Dantas

Source & license

This open-source MCP server is cataloged on AgentStack and links to its original source — we do not rehost the code.

Install and usage instructions live in the source repository linked above.

Reviews

No reviews yet — be the first.

Versions

  • v0.1.0 Imported from the upstream source.