FAQ · 10 questions
Questions, answered plainly.
Who runs Qualixar, what the tools cost, and how to check anything we claim.
What is Qualixar?
Qualixar is the independent AI Reliability Engineering research initiative of Varun Pratap Bhardwaj. It publishes 17 open-source tools and 10 arXiv preprints on making AI agents testable, bounded and auditable.
Is Qualixar a company?
No. It is an independent research initiative. The tools are open source; commercial licences are available on request for teams that need them.
Are the papers peer-reviewed?
No. They are preprints on arXiv. Each paper page links to the arXiv abstract, the PDF and the code, so you can read the method and check the results yourself.
Are the tools free to use?
Yes. Every tool is published with its source code. The licences differ by tool (AGPL-3.0, AGPL-3.0 or commercial, AGPL-3.0-or-later, Apache-2.0, Elastic-2.0, FSL-1.1-ALv2, MIT), and each repository states its own. Read the licence before you ship a tool inside a product.
How do I check a number on this site?
Every figure on a product page or the home page links to the paper, repository or package registry it comes from. Product pages have a Verify panel listing those sources. Figures a project reports about itself, such as test counts, are labelled that way.
Does SuperLocalMemory send my data to the cloud?
SuperLocalMemory is local-first: the core memory data stays on infrastructure you control, and its zero-LLM modes retrieve without calling a model. Which modes you run is your choice; the product page and README explain each one.
Where should I start?
Pick the problem you have: memory, evaluation, loops, security or contracts. Each topic hub lists the tools, papers and posts on it, with a reading path that keeps your place in this browser.
What does a term like vacuous gate or stochastic verdict mean?
The glossary defines the terms used across the tools and papers in plain words, and links each one to its source.
How do I report a bug or ask for a feature?
Open an issue on the tool's GitHub repository. Each product page links to its repository. For anything else, use the contact page.
How do I know what changed in a tool?
The changelog collects the latest releases across the toolkit, newest first, and each product page shows its own recent releases.
Still have a question?
The AI Reliability Engineering guide answers the bigger ones, and the contact page reaches Varun directly.