Skip to content
The DSA Woodshed

The DSA Woodshed

Woodshedding for the whiteboard.

Musicians woodshed: practice the hard passages alone until the performance holds under pressure. This is that room for technical interviews: real problems, real code, and a session shaped around the time you have.

Choose study, implementation, tests first, or a conversation before you begin. Write your reasoning in ordinary # comments and docstrings. The Python track ships 72 algorithm exercises, a printable booklet, and reference sheets.

Choose a session Read the booklet How it works

Basic practice needs no repository API key or private-service credentials.

The default

Start with ordinary comments

Choose study or implementation, then write your reasoning in ordinary comments. There are no required prefixes or labels.

Choose an available problem and open its isolated source and test files. Personal interviewer integrations can guide the same loop with optional labels.

One clean loop

How a rep works

Choose an available capability and a time budget. For an implementation session, use these four movements.

  1. Write the thinking

    Restate the problem, work an example, and name an approach.

  2. Build and test

    Implement in your private, gitignored workspace and add focused cases to your test file.

  3. Reconcile the receipt

    Run the focused tests; a failed, stale, or missing run asks you to trace, revise, and rerun.

  4. Keep one correction

    Finish with one correction to carry into the next rep.

You choose when to test.

Starting or resuming a session does not run tests. Study keeps the committed solution and reference tests intact; implementation uses an isolated candidate workspace. Start focused tests when you choose to check your work.

The point

Practice explaining your work

Solving the problem is one skill. Explaining your examples, assumptions, and corrections is another. Use a talk session to form a plan, a board session to narrate within your chosen budget, or a mock session to rehearse the full conversation.

Your comments can be the whiteboard. Keep one concrete correction, then stop or choose another session. The practice evidence guide explains the method and its limits.

Review with a purpose

Library

See the full library

Study a committed solution and its reference tests before a rep, review a complete solution after a gap, or take a printable sheet away from the screen.

Match the surface to the skill

Choose study when you want the committed source and reference tests open without starting a rep. Choose a talk session to slow down and form a plan, or a board session to rehearse explaining within the time you have. Use your own clock for a timed rep; the session command records your budget without starting a timer.

Prefer local VS Code or no agent at all? The same workflow runs through the documented just recipes.