Quick Start Guide

Get up and running with DynaLab.ai in under 5 minutes. This guide walks you through account creation, onboarding, and taking your first assessment.

Create Your Account

DynaLab.ai supports three authentication methods. Choose whichever is most convenient for you:

  • GitHub OAuth — Sign in with your GitHub account. Fastest option if you already have one.
  • Google OAuth — Sign in with your Google account.
  • Email OTP — Enter your email address and verify with a one-time code. No third-party account needed.

For hiring teams

If you received an invitation link from an employer, click that link first. It will guide you through account creation and automatically connect you to the team assessment.

Onboarding Flow

After creating your account, you'll go through a brief onboarding to personalize your experience:

  1. Welcome — A quick introduction to what DynaLab.ai measures and how it works.
  2. Role Questions — Optional questions about your experience level and focus areas. Helps us recommend relevant tasks.
  3. Scoring Intro — Brief explanation of the 8 scoring dimensions so you know what's being measured.
  4. Diagnostic Assessment — A short 4-minute, 6-8 question assessment to establish your baseline skill level. This personalizes your learning path recommendations.

Diagnostic is optional

You can skip the diagnostic and jump straight to tasks, but completing it unlocks personalized task recommendations on your dashboard.

Your First Assessment

Choosing a Task

Browse the task library from the dashboard or the Tasks page. Each task shows:

  • Difficulty level (beginner, intermediate, advanced)
  • Category (bug fix, code review, debugging, performance, etc.)
  • Language and framework
  • Time limit (typically 30 minutes)
  • Whether it's free or requires a paid plan

18 tasks are free for all users. Start with a beginner-level task like Task 001: Fix Intermittent 500 Errors to get comfortable with the IDE.

Taking the Assessment

  1. Click "Start" on a task to launch the IDE. A sandboxed coding environment loads with the task's codebase.
  2. Read the brief — The task panel on the right describes the problem, expected behavior, and any hints.
  3. Explore before acting — Browse files, read logs, understand the codebase before jumping to a fix. This is one of the most heavily weighted scoring signals.
  4. Use the AI chat — Ask the AI assistant questions, request code changes, and iterate. How you interact with the AI is what gets scored.
  5. Run tests — Use the terminal to run the project's test suite and verify your changes work.
  6. Submit — Click the Submit button when you're satisfied with your work (or when time runs out).

Understanding Your Scorecard

After submission, your session is scored across 7 calibrated dimensions. The scorecard typically takes 30-60 seconds to generate and includes:

  • Overall score (0-100) and letter grade (S/A/B/C/D/F)
  • Dimension scores — How you performed on each of the 8 scoring dimensions
  • Strengths and weaknesses — Key areas where you excelled or need improvement
  • Key moments — Specific timestamped events from your session (e.g., catching a hallucination, running tests after AI changes)
  • Session replay — A full timeline of your session that you can replay

See the Scoring System docs for a detailed breakdown of how each dimension is calculated.

Next Steps

Next
IDE & Assessments