bmad-code-org__bmad-method/docs/build/test-completed-work.md at main

mirror of https://github.com/bmad-code-org/BMAD-METHOD.git synced 2026-09-19 08:11:52 +08:00

Files

T

Alex Verkhovsky 3bc2271260 docs: add Customize and Extend chapter (#2797 )

Create the Customize and Extend chapter after Existing Codebases with five
pages under docs/customize/: Customize BMad, Adopt BMad Across a Team,
Add Modules, Use Web Bundles, Run Multi-Agent Discussions.

Retire eight English pages with redirects: how-to/customize-bmad,
explanation/named-agents, how-to/expand-bmad-for-your-org,
how-to/install-custom-modules, reference/modules, how-to/use-web-bundles,
explanation/web-bundles, explanation/party-mode.

Corrections against shipped sources: the installer has no community
catalog browser (one custom/community prompt with an unverified-module
warning); bmad-prd's checklist key is validation_checklist_template; the
brief template ships under assets/; doc_standards passes run in declared
order; the web-bundle shelf personas now match web-bundles/bundles.json.

2026-08-30 04:22:29 -06:00

4.8 KiB

Raw Permalink Blame History

title, description, sidebar

title

description

sidebar

Test Completed Work

Choose a generate skill after implementation — simple coverage with bmad-qa-generate-e2e-tests, or heavier coverage with bmad-testarch-automate.

order
4

After a change is implemented, decide whether it needs more automated coverage and which generate skill should produce it. Both bmad-qa-generate-e2e-tests and bmad-testarch-automate generate tests from code that already exists. The built-in skill stays simple. Automate is the heavier generate: fixtures, more test levels, and knowledge-base patterns. See how the built-in skill runs.

This is generated coverage of finished work. It is not code review, and it is not the manual observations in Walk Through a Change.

Which Path?

Factor	`bmad-qa-generate-e2e-tests`	`bmad-testarch-automate`
Best for	Simple coverage of implemented features	Heavier coverage of the same kind of work
Setup	Included with BMM	Install the TEA module
Approach	Generate from the code that exists	Same, standalone; optional test design improves the run
What it covers	API and E2E; happy path plus a few errors	API, E2E, fixtures, more patterns; optional component tests

:::tip[Start with built-in QA] Most projects should start with bmad-qa-generate-e2e-tests. Use bmad-testarch-automate when you want the heavier generate of the same kind of work. :::

Other TEA skills — test design, trace, ATDD, test review, NFRs, and gates — are available. They are not the default generate path.

Run `bmad-qa-generate-e2e-tests`

Open a fresh chat and name the skill. You can say what to test before, with, or after the command — a feature, a directory, or "discover what is untested."

/bmad-qa-generate-e2e-tests

/bmad-qa-generate-e2e-tests Create API and E2E tests for the login flow.

It uses whatever test framework the project already has. If there is none, it looks at the stack and suggests one.

What a run does

Detect the test framework — scans dependencies and existing tests (Playwright, Jest, Vitest, Cypress, and similar).
Identify features — asks what to test, or auto-discovers features in the codebase.
Generate API tests when there are endpoints — status codes, response shape, happy path, and one or two error cases.
Generate E2E tests when there is a UI — user workflows with semantic locators (roles, labels, text) and visible-outcome assertions.
Run the tests and fix failures immediately.
Write a summary of what was generated and what is still uncovered.

Generated tests stay simple on purpose: standard framework APIs, independent cases, no hardcoded waits, descriptions that read as feature documentation.

What You Get

Test files under the project's tests/ directory
A test summary at tests/test-summary.md in your implementation artifacts directory
Tests that were run once in this session and made to pass

Limits

bmad-qa-generate-e2e-tests generates tests only. It does not review the implementation — that is bmad-build during the run, or bmad-code-review if you want another pass.

Happy path plus a few critical errors is the ceiling, and it does not compose complex fixtures. More edge cases are follow-up work, or a reason to use Automate.

When to Use TEA

Install the TEA module when you want bmad-testarch-automate — heavier generate of the same kind of work. Automate still generates from existing code and can run standalone.

bmad-testarch-test-design is optional before Automate. It improves the run; it is not a prerequisite.

bmad-testarch-trace is optional after generation, to check coverage. It is not part of generate.

ATDD is for features that do not exist yet. Test review, NFR assessment, and release gates are available when you need them. They are not required to generate tests.

TEA is a separate module. Its current workflows, commands, and setup live in the TEA documentation. Install it with the rest of BMad; see Add Modules for how modules are selected.

Where It Fits

bmad-build implements a change and, if a suite already exists, aims to leave those tests passing. This page is the next testing decision: generate additional coverage for that finished work.

You can run built-in QA after one change. You do not have to wait for an epic to finish. A typical sequence is implement with bmad-build, optionally walk through the result, then generate coverage here. After a whole epic, bmad-retrospective is a different check — it judges the epic against its spec, not the test suite.

4.8 KiB Raw Permalink Blame History