AI-powered deployment verification, secret scanning, incident analysis, infrastructure auditing, runbook generation, and cost optimization. Catches leaked AWS keys, GitHub tokens, database URLs, and private keys. Free tier runs locally with zero data transmission. Every action is governed and receipted.

For AI Assistants

This MCP server returns structured JSON for seamless integration:

Secret detection with pattern type (AWS, GitHub, JWT, private key, DB URL)
Incident severity classification with root cause analysis
Governance receipt on every call (audit-ready)
Chain-ready: deploy_check pre-deploy -> secret_scan for leaked credentials -> infra_audit for config drift -> incident_analyze when issues arise -> runbook_generate for response procedures -> cost_optimize for spend reduction

Composable with any MCP client: Claude Code, Cursor, VS Code, ChatGPT Desktop, Windsurf.

Install

npx dingdawg-devops-agent

Claude Code

claude mcp add devops -- npx dingdawg-devops-agent

Cursor

Add to .cursor/mcp.json:

{"mcpServers": {"devops": {"command": "npx", "args": ["dingdawg-devops-agent"], "env": {"DINGDAWG_API_KEY": "your-key"}}}}

Full Stack (all 13 agents)

npx dingdawg-setup

Tools

Tool	Free Tier	Paid Tier
`deploy_check`	10/day, basic deploy verification	Unlimited, LLM-powered with rollback recommendations
`secret_scan`	10/day, 7 secret patterns (AWS, GitHub, JWT, etc.)	Unlimited, 50+ patterns with rotation guidance
`incident_analyze`	5/day, basic log analysis	Unlimited, AI-powered root cause analysis
`infra_audit`	5/day, config checklist	Unlimited, drift detection with remediation
`runbook_generate`	3/day, template-based	Unlimited, AI-generated context-aware runbooks
`cost_optimize`	3/day, basic spend analysis	Unlimited, AI-powered with savings projections

Pricing

Free: 10 checks/day, basic analysis
Pro: $49/mo, 100 calls/day, AI-powered deep analysis
Pay-as-you-go: $0.25/call, no commitment

Get API key: https://dingdawg.com/developers

Governed

Every call is receipted and auditable. Secret scans reference specific pattern types with zero false-positive-tolerant matching. Incident analyses include timeline reconstruction. Infrastructure audits reference CIS benchmark controls.