Portfolio Pro: Proof of Work for developer hiring

JavaScript is required to use this application. Please enable JavaScript in your browser settings to continue.

Send a candidate one JavaScript challenge and get back their code, the tests it passed, and how the code arrived. Portfolio Pro also offers a catalog of guided AI development tracks and projects.

Portfolio ProShip, don't watch

Loading tracks…

Portfolio Pro

Loading page…

Premium track: full modules and projects with a plan

View plans

Learning track · Advanced

AI Agents & Autonomous Systems

Build autonomous AI agents that can reason, plan, and execute complex tasks. Learn tool calling, MCP servers, multi-agent orchestration, evaluation harnesses, and how to deploy agents to production.

Vercel AI SDK
Model Context Protocol (MCP)
TypeScript
OpenAI
LangGraph-style Orchestration

Modules: 5
Lessons: 26
Projects: 12
Pace: 8-10 weeks

Curriculum

Modules

5 modules · 26 lessons · 12 projects

01
Agent Fundamentals
Loops, memory, planning, and the minimum architecture for a reliable tool-using agent.
6 lessons
What Makes An Agent
The Agent Loop and State
Function Calling and Structured Outputs
Vercel AI Sdk Agent Basics
+2 more
2 projects
Agent Scratchpad Notebook
Tool-Calling Agent
Premium
02
Tool Use Patterns
Beyond a single tool call: schema design at scale, parallel vs. sequential execution, retrieval as a tool, retry strategies, and streaming tool activity to a UI.
6 lessons
Designing Tool Schemas with Zod
Parallel Vs Sequential Tool Calls
Retrieval and RAG as a Tool
Error Handling and Retries
+2 more
3 projects
AI Code Review Bot
AI Meeting Summarizer
CLI Tool Builder
Premium
03
MCP Integration
Build and deploy Model Context Protocol servers agents can plug into Cursor, Claude, and your apps.
5 lessons
MCP Mental Model
Stdio Servers
Streamable HTTP
Connecting MCP Clients to Agents
+1 more
2 projects
Custom MCP Server
MCP Server CLI
Premium
04
Multi-Agent Systems
Specialize agents, route work, and keep orchestration debuggable.
5 lessons
When to Multi-Agent
Supervisor Pattern
Graph Orchestration
Parallel Agent Execution and Aggregation
+1 more
2 projects
Multi-Agent Research Team
Support Ticket Triage Swarm
Premium
05
Production Agents
Ship it: build an eval harness that catches regressions before users do, add observability and cost tracking, deploy for real, and lock down rate limits and abuse prevention.
4 lessons
Evaluation Harnesses for Agents
Observability Tracing and Cost Tracking
Deploying Agents to Production
Safety Rate Limits and Abuse Prevention
3 projects
Agent Eval Harness
Observed Agent Dashboard
Production Agent API
Premium

Outcomes

What you ship

Ship 12+ portfolio projects from this track
Build autonomous AI agents with tool calling and structured outputs
Design and ship custom MCP servers (stdio + Streamable HTTP)
Orchestrate multi-agent systems with supervisor and graph patterns
Build eval harnesses that catch regressions before users do
Deploy, observe, and rate-limit agents in production

Premium track: full modules and projects with a plan

View plans

Learning track · Advanced

AI Agents & Autonomous Systems

Build autonomous AI agents that can reason, plan, and execute complex tasks. Learn tool calling, MCP servers, multi-agent orchestration, evaluation harnesses, and how to deploy agents to production.

Vercel AI SDK
Model Context Protocol (MCP)
TypeScript
OpenAI
LangGraph-style Orchestration

Modules: 5
Lessons: 26
Projects: 12
Pace: 8-10 weeks

Curriculum

Modules

5 modules · 26 lessons · 12 projects

01
Agent Fundamentals
Loops, memory, planning, and the minimum architecture for a reliable tool-using agent.
6 lessons
What Makes An Agent
The Agent Loop and State
Function Calling and Structured Outputs
Vercel AI Sdk Agent Basics
+2 more
2 projects
Agent Scratchpad Notebook
Tool-Calling Agent
Premium
02
Tool Use Patterns
Beyond a single tool call: schema design at scale, parallel vs. sequential execution, retrieval as a tool, retry strategies, and streaming tool activity to a UI.
6 lessons
Designing Tool Schemas with Zod
Parallel Vs Sequential Tool Calls
Retrieval and RAG as a Tool
Error Handling and Retries
+2 more
3 projects
AI Code Review Bot
AI Meeting Summarizer
CLI Tool Builder
Premium
03
MCP Integration
Build and deploy Model Context Protocol servers agents can plug into Cursor, Claude, and your apps.
5 lessons
MCP Mental Model
Stdio Servers
Streamable HTTP
Connecting MCP Clients to Agents
+1 more
2 projects
Custom MCP Server
MCP Server CLI
Premium
04
Multi-Agent Systems
Specialize agents, route work, and keep orchestration debuggable.
5 lessons
When to Multi-Agent
Supervisor Pattern
Graph Orchestration
Parallel Agent Execution and Aggregation
+1 more
2 projects
Multi-Agent Research Team
Support Ticket Triage Swarm
Premium
05
Production Agents
Ship it: build an eval harness that catches regressions before users do, add observability and cost tracking, deploy for real, and lock down rate limits and abuse prevention.
4 lessons
Evaluation Harnesses for Agents
Observability Tracing and Cost Tracking
Deploying Agents to Production
Safety Rate Limits and Abuse Prevention
3 projects
Agent Eval Harness
Observed Agent Dashboard
Production Agent API
Premium

Outcomes

What you ship

Ship 12+ portfolio projects from this track
Build autonomous AI agents with tool calling and structured outputs
Design and ship custom MCP servers (stdio + Streamable HTTP)
Orchestrate multi-agent systems with supervisor and graph patterns
Build eval harnesses that catch regressions before users do
Deploy, observe, and rate-limit agents in production

AI Agents & Autonomous Systems

Modules

Agent Fundamentals

Tool Use Patterns

MCP Integration

Multi-Agent Systems

Production Agents

What you ship

AI Agents & Autonomous Systems

Modules

Agent Fundamentals

Tool Use Patterns

MCP Integration

Multi-Agent Systems

Production Agents

What you ship