Changelog
What's new in TestLookup
Everything built so far is heading for the first official release, 0.1.0. Here's what it contains, what changed most recently and what to check before you update an install built from the main branch.
Release status
TestLookup hasn't published an official release yet. The repository's VERSION file says 0.1.0, and the changelog gathers all current work under an unreleased 0.1.0 heading. It calls this the first public release.
Until 0.1.0 is released, you run TestLookup from the source code, or from the container images built for every change to the main branch. The documentation on this site describes that development version and follows the main branch.
TestLookup is early, pre-1.0 software maintained by volunteers, so expect some bugs. Before version 1.0, even a minor version can include changes that break an existing setup.
No releases yet
GitHub showed no releases and no tags for the repository. A version number in the changelog or a package file does not mean that version has been released.
Checked on .
Newest changes
The latest work added to the code.
Data
Export a project with its run history
An export can now include finished runs from the last 30, 90 or 365 days, or all of them, up to 2,000 runs. It comes as one signed archive, and you see an estimate first.
Data
Imports that pick up where they stopped
Archives up to 2 GB upload in 32 MB parts, so an interrupted upload can resume. An import always creates a new project, and it starts no analysis and sends no notifications.
AI
More accurate run summaries
When a summary is written from a given list of failures, it now names those failures instead of describing the saved run.
What 0.1.0 contains
The main themes the changelog lists for the first release.
Test results
Every common report format
JUnit, TestNG, Robot Framework, NUnit 3, TRX, xUnit, Playwright, pytest, Cypress, Cucumber and Allure reports, plus live results from 11 test runners while tests run.
Runs and failures
From a failed build to its cause
Each test run gets a page with its tests, analysis, code changes, evidence and timings. You can compare runs, group similar failures, score how flaky each test is, and send defects to Jira.
Releases
Release decisions
A release check that says whether a release is ready, using versioned rules. It adds release phases, a calendar and rules for which test runs count. Every override is logged, and signed evidence bundles are ready for auditors.
AI
AI you can check
TestLookup sorts each failure using rules, machine learning or an AI language model, and shows where each result came from. Analyses are versioned. There's also deep investigation, Ask AI, a knowledge base and quality checks for the AI.
Test management
Test cases beside results
Manage test cases, plans, strategies and reviews next to your results. You also get versions, saved views, a health check of your test library, CSV or Excel import and requirement links.
Operations
Run it on your own servers
Install with Docker Compose, Kubernetes, K3s or OpenShift, including on a network with no internet access. Includes single sign-on, automatic user setup, two-step sign-in, retention holds, backups and upgrades.
Updating an install built from main
If you already run a copy of TestLookup built from the main branch, read these changes and the upgrade notes before you update it.
Storage
New bundled message queue and file storage
TestLookup now ships with Valkey as its message queue and versitygw as its file storage. Both start empty, and stored files are copied over once from the previous store.
AI
New default local model
The default local AI model is now qwen2.5:7b-instruct. ChromaDB data moves to a new location, so the search-by-meaning index is rebuilt.
Security
No default admin password
A new install creates a random admin password, shows it once and makes you change it the first time you sign in.
Follow along
The full record is in the TestLookup repository on GitHub.