{
  "markdown": "# AI Integration Architect — Claude Code Plugin\n\nA Claude Code plugin that helps teams plan, design, and scaffold AI integration into enterprise systems.\n\n**46% of enterprises say integrating AI into existing systems is their #1 challenge.** This plugin acts as a senior integration architect — it assesses your system landscape, recommends architecture patterns, generates production-ready code scaffolds, and produces deployment plans with security baked in.\n\n## Installation\n\n### Option 1: One-command install via marketplace (recommended)\n\n```bash\nclaude plugin marketplace add pfbarros2/ai-integration-architect-plugin\nclaude plugin install ai-integration-architect\n```\n\n### Option 2: Install directly from the skill repo\n\n```bash\nclaude plugin marketplace add pfbarros2/ai-integration-architect\nclaude plugin install ai-integration-architect\n```\n\n### Option 3: Local install\n\n```bash\ngit clone https://github.com/pfbarros2/ai-integration-architect-plugin.git\nclaude --plugin-dir ./ai-integration-architect-plugin\n```\n\n## What it does\n\n| Phase | Output |\n|-------|--------|\n| **Assess** | Maps your current systems (APIs, databases, SaaS tools) and identifies where AI adds the most value |\n| **Architect** | Recommends integration patterns (MCP servers, API gateways, RAG pipelines, event-driven agents) with trade-off analysis |\n| **Scaffold** | Generates working starter code — MCP servers, API connectors, RAG pipelines, middleware — with auth, rate limiting, error handling, and tests |\n| **Deploy** | Produces deployment configs, monitoring setup, security checklists, cost estimates, and phased rollout plans |\n\n## Example prompts\n\nOnce installed, just describe your integration challenge:\n\n> \"We run Shopify Plus with Salesforce CRM and Zendesk for support. Our CEO wants AI to help reduce support response times. Where do I start?\"\n\n> \"I need an MCP server that connects Claude to our internal REST API. It uses OAuth2, has /employees, /projects, and /timesheets endpoints.\"\n\n> \"We have 5000 pages of docs across Confluence, Google Docs, and Notion. Engineers waste hours searching. I want a RAG pipeline so our AI assistant can answer questions from any internal doc.\"\n\n> \"Design an event-driven architecture where an AI agent triages incoming support tickets in Zendesk, looks up customer history in Salesforce, and drafts responses.\"\n\n## Supported integration patterns\n\n- **MCP Servers** — Connect Claude to any API, database, or internal tool (TypeScript + Python templates)\n- **API Gateways** — Expose AI as a service for your existing systems to call, with model routing (Haiku/Sonnet/Opus by complexity)\n- **RAG Pipelines** — Ingest enterprise knowledge from Confluence, SharePoint, Notion, Google Docs; chunk, embed, retrieve with reranking\n- **Event-Driven Agents** — AI that responds to system events (new ticket, alert, deployment) with human-in-the-loop approval gates\n- **Hybrid Architectures** — Combine patterns for complex workflows\n\n## Security built in\n\nEvery output includes enterprise security by default:\n\n- Least-privilege service accounts and scoped API keys\n- Audit logging for all AI-initiated actions\n- Data classification guidance (Public → Internal → Confidential → Restricted)\n- Compliance checklists for SOC 2, GDPR, and HIPAA\n- Prompt injection mitigation strategies\n\n## Plugin structure\n\n```\nai-integration-architect-plugin/\n├── .claude-plugin/\n│   └── plugin.json\n├── skills/\n│   └── ai-integration-architect/\n│       ├── SKILL.md\n│       └── references/\n│           ├── patterns.md\n│           ├── scaffolds.md\n│           ├── security.md\n│           └── deployment.md\n├── marketplace.json\n└── README.md\n```\n\n## License\n\nMIT\n",
  "bytes": 3673,
  "sha": "cd2d24b40ae81b5d3c4745d807536b28d905da4d2b61d98c44f30194111ec5da",
  "repo_slug": "pfbarros2/ai-integration-architect-plugin",
  "fonte": "repo",
  "truncated": false,
  "api": "https://agentalog.com/api/listings/plg_pfbarros2_ai_integration_architect_plugi_bf98762b/readme"
}