OPENSPEC · HANDS-ON LABS

Lab bundles

Every lab ships three folders: starter (where you begin), solution (one correct answer), and expected_output (what a finished run looks like). Take the bundle that matches your stack.

← Back to the OpenSpec course

How to use these

Unzip, then work lab by lab in order — each one assumes the OpenSpec directory the previous lab created. Start from starter/ and read that lab's README.md; only open solution/ once you have your own answer, because comparing two answers teaches more than reading one.

You need the OpenSpec CLI and an AI coding assistant. The labs were written against Claude Code, but the workflow is assistant-agnostic:

npx openspec@latest init --tools claude
npx openspec@latest validate
Python · pytest

OpenSpec Labs — Python

The main track. Nine labs from initialising OpenSpec in a repo through spec authoring, change proposals, implementation, team CI, security guardrails and a capstone.

01 setup & init02 name normalizer03 spec authoring 04 proposal05 implement06 team CI 07 security guardrails08 governance health09 capstone
⬇ Download bundle 105 KB · 116 files
Java · Spring Boot · Maven

OpenSpec Labs — Java

The same nine-lab arc on a Spring Boot service. Lab 06 is worth the trip on its own: a CI workflow that fails the build when a spec scenario has no matching test.

01 setup & init02 name normalizer03 spec authoring 04 proposal05 implement06 team CI 07 security guardrails08 governance health09 capstone
⬇ Download bundle 141 KB · 168 files
Python · PySpark · Airflow · Great Expectations

OpenSpec Labs — Data Engineering

Ten labs building a bronze / silver / gold pipeline specified before it is built, with data-quality expectations as acceptance criteria and a handoff capstone. Runs locally — no cloud account needed.

01 openspec setup02 planning03 environment 04 bronze ingest05 silver normalize06 gold risk profile 07 airflow08 great expectations09 testing & CI10 capstone
⬇ Download bundle 711 KB · 704 files