Docs testing guides: keep your code examples working
How to test the code in your documentation, catch drift before readers do, and pick the right checks for each layer of your docs. Each guide uses real findings from our own runs, never made-up numbers.
Guide · start hereDocs Testing: How to Test the Code Samples in Your Docs
Why code examples in documentation break, the ways teams test them (doctest, docs-as-tests, CI runners) and how to run every sample on each release.
ComparisonVale vs markdownlint: What Each Docs Linter Catches
Vale checks prose style; markdownlint checks Markdown structure. What each catches, sample configs, running both in CI and what neither one tests.
ComparisonDoc Detective vs DocsAlive: Testing the Code in Your Docs
Doc Detective is a free, open-source docs-as-tests framework; DocsAlive is a hosted docs CI for code examples. Features, price, setup and when to use each.
ComparisonPromptless Alternative: Promptless vs DocsAlive Compared
Promptless drafts docs updates from what you ship; DocsAlive runs the code examples you've published. Features, pricing, security and when to use each.
GuideBest Free Tools to Test Code Examples in Documentation
Free, open-source tools that run the code examples in your docs: rustdoc, Go examples, doctest, pytest plugins, Sphinx doctest, Doc Detective and byexample.
GuideTesting SDK Docs Before Every Release: A Release-Gate Guide
Test SDK and client-library docs against the package you ship: build the wheel or npm tarball, install it clean, run every example, and gate the release in CI.
GuideDocumentation Drift: How to Detect It Before Users Do
What documentation drift is, why code examples drift first, and practical ways to detect it: ownership, release checks, link and execution tests.
Case studyDuckDB fixed a broken docs example that DocsAlive caught
Two record-batch examples crashed with StopIteration. We reported a tested fix and DuckDB merged a fix seven days later.
ResultsReal runs on three open-source docs sites
Every likely docs bug the engine flagged, which ones we confirmed, and where it got it wrong.
How we write these.
- Real numbers only. Every figure about broken docs comes from our own published runs on the results page. We don't quote industry statistics we can't source.
- Checked against current docs. Config keys, flags and action versions are verified against each tool's own documentation on the date shown, and the page lists what was checked.
- Credit where it's due. We link to the open-source tools we compare and say plainly when one of them is the better fit.
- We sell DocsAlive. Read the comparisons with that in mind. We keep the product to one clearly labelled section per page.
Test your docs, not just your code.
Free health check for public repositories. Real report, no signup.