Open-source platform to learn about AI Agents through structured roadmaps and curated content.
Live: https://agent-foundry-theta.vercel.app
Agent Foundry is a free, open-source learning platform — built specifically for developers who want to master AI agent frameworks. It provides structured roadmaps with hands-on code examples covering LangChain, CrewAI, OpenAI Agents SDK, LangGraph, and more.
| Roadmap | Topics | Description |
|---|---|---|
| LangChain | 12 | Chains, agents, tools, memory, RAG, and production patterns |
| CrewAI | 8 | Multi-agent orchestration, roles, tasks, and processes |
| OpenAI Agents SDK | 8 | Function calling, handoffs, guardrails, and the agent loop |
| LangGraph | 8 | Stateful agent graphs, nodes, edges, and persistence |
Each roadmap is organized into sections with topics tagged by difficulty level (Beginner, Intermediate, Advanced).
- Framework: Next.js 16 (App Router, React Server Components)
- Styling: Tailwind CSS v4 + shadcn/ui
- Icons: Lucide React
- Theming: next-themes (dark/light mode)
- Content: MDX files rendered via next-mdx-remote
- Deployment: Vercel
agent-foundry/
├── content/
│ └── roadmaps/
│ ├── langchain/ # meta.json + topic .mdx files
│ ├── crewai/
│ ├── openai-agents-sdk/
│ └── langgraph/
├── src/
│ ├── app/
│ │ ├── page.tsx # Landing page
│ │ └── roadmaps/
│ │ ├── page.tsx # Roadmaps index
│ │ └── [slug]/
│ │ ├── page.tsx # Roadmap detail (structured sheet)
│ │ └── [topicSlug]/
│ │ └── page.tsx # Topic page (MDX content)
│ ├── components/
│ │ ├── ui/ # shadcn/ui components
│ │ ├── navbar.tsx
│ │ ├── footer.tsx
│ │ ├── roadmap-card.tsx
│ │ ├── topic-row.tsx
│ │ ├── difficulty-badge.tsx
│ │ ├── mdx-content.tsx
│ │ └── topic-nav.tsx
│ └── lib/
│ ├── content.ts # Filesystem content loading utilities
│ ├── types.ts # TypeScript types
│ └── utils.ts # shadcn utility
└── package.json
- Node.js 18+
- npm
git clone https://github.com/ishandutta0098/agent-foundry.git
cd agent-foundry
npm installnpm run devOpen http://localhost:3000 to view the app.
npm run buildAll content lives in the content/roadmaps/ directory. To add a new roadmap:
- Create a new folder under
content/roadmaps/(e.g.,content/roadmaps/autogen/) - Add a
meta.jsondefining the roadmap structure:
{
"title": "AutoGen",
"description": "Learn multi-agent conversations with AutoGen",
"icon": "bot",
"sections": [
{
"title": "Getting Started",
"topics": [
{ "slug": "introduction", "title": "What is AutoGen?", "difficulty": "beginner" }
]
}
]
}- Create
.mdxfiles for each topic slug (e.g.,introduction.mdx) - The new roadmap appears automatically on the site
The project is deployed on Vercel with automatic deployments on every push to main.
- Fork this repository
- Go to vercel.com/new and import your forked repo
- Vercel auto-detects Next.js — no configuration needed
- Click Deploy and your site will be live in under a minute
Alternatively, deploy via the Vercel CLI:
npm install -g vercel
vercel login
vercel --prodEvery subsequent git push to main triggers an automatic production deployment. Pull requests get preview deployments with unique URLs for testing changes before merging.
All pages are statically generated at build time (SSG), resulting in fast load times and zero server costs on Vercel's free tier:
/— Static landing page/roadmaps— Static roadmaps index/roadmaps/[slug]— Pre-rendered per roadmap viagenerateStaticParams/roadmaps/[slug]/[topicSlug]— Pre-rendered per topic viagenerateStaticParams
Contributions are welcome! You can:
- Add new topic content (MDX files) to existing roadmaps
- Create entirely new roadmaps for other AI agent frameworks
- Improve existing content, fix errors, or enhance explanations
- Suggest features or report issues via GitHub Issues
This project is licensed under the GNU General Public License v3.0.