Morluto on REA: The Anonymous Founder's First Interview

A project can be both a technical tool and a question people suddenly realize they are allowed to ask. In Morluto's first interview with Matt Van Horn on The /last30days Podcast, the pseudonymous founder of Reverse Engineer Anything (REA) explains how a small experiment became a viral open-source project. The interesting part is not just the GitHub chart. It is Morluto's belief that agents let more people inspect a system, understand why it works the way it does and build a better answer to their own problem.

Watch the First Interview

The founder speaks under the name Morluto. This article does not try to identify them. The origin story and opinions below are attributed to the interview; REA's capabilities, setup and use boundaries are checked against the official repository as it stood on 29 October 2026.

From an Open-Source PR to REA

Morluto says one of their first open-source contributions was a fix to Garry Tan's GStack. They noticed Matt's frequent, high-quality contributions and were surprised to learn how much of his workflow was agent-assisted. That experience made open-source contribution feel more accessible: an agent could help with unfamiliar code, while a person still had to care about the change and take responsibility for it.

The immediate spark for REA came later. Morluto tried a newly launched agent workspace, mistook a helpful team member for a bot, and started reporting bugs. Wanting to understand the app more deeply, they connected analysis tools through MCP and watched an agent inspect a closed-source application. The first experiment was less a grand plan to replace software than a mix of curiosity, a desire to help the team, and the realization that a tool normally reserved for specialists might be made usable through natural-language investigation. Morluto credits a collaborator with developing important parts of what REA became.

That account matters because the project did not begin with a market-size argument or a clone factory. It began with a concrete question about software the founder was already using. Morluto's advice to newcomers is implicit in the story: start with a real behavior you want to understand, not the largest app you can name.

What REA Actually Does Today

The current README describes REA as an MCP server and matching agent workflow for investigating app behavior, files and binaries. An agent can call analysis tools, inspect returned evidence and use what it learns to explain a feature or write and test an original implementation. REA also exposes command-line workflows. That is more useful than the slogan "clone anything," and it leaves room for incomplete evidence and a failed investigation.

The repository lists many more target types, including .NET assemblies, Android artifacts and firmware, each with different prerequisites. Its setup guide is the right place to check those details. REA says its analysis runs locally, but an agent's model provider may still receive the resulting context; local analysis is not the same as a fully offline workflow.

359K in the Headline, Nearly 300K on Camera

Three numbers describe three different moments. Matt says the repository is approaching 300,000 stars at the time of recording. The video's title says 359,000 in two weeks. On 29 October 2026, the official GitHub page displayed about 405,600 stars. None is a permanent count, and none proves how many people installed, used or kept the tool.

Morluto pushes back on treating stars as a product-quality score. A repository can attract attention through timing and marketing; a quiet one can solve a difficult problem well. For a builder, better questions are whether the supported target matches your task, whether the evidence is inspectable, whether the workflow stays within your permissions, and whether a reproduced feature actually helps a user.

Is This the Death of Software?

Morluto rejects that framing. In the interview, they describe a small irritation in software they use, a control that could not be changed, as the sort of limitation people once had to accept. The new possibility is to investigate how the behavior works and make an alternative that suits a different need. More people may be able to modify or rebuild software, but choosing what to make, designing it well and maintaining it are still substantial work.

This is the strongest version of the interview's argument: lowering the cost of understanding an interface can increase the demand for good product judgment. It does not follow that every paid app is obsolete. Products still need reliable data, distribution, support, security, legal rights and a reason for customers to trust the result. REA makes one part of the process more approachable; it does not supply those missing parts.

The Boundary Between Learning and Copying

The conversation briefly touches on misuse, including unauthorized access and attempts to get around licenses. The project's own responsible-use disclaimer is explicit: it does not support bypassing DRM or license checks, breaking into systems, or copying proprietary code and protected assets. REA's MIT license covers the REA project; it does not grant rights to the software being examined.

A defensible first use is an app you own, an artifact you have permission to inspect, or a narrowly scoped interoperability or research task with the appropriate rights. Keep the analysis question and the new product's design separate. Evidence from a decompiler can explain behavior, but it is not a license to ship someone else's code or visual identity. REA's disclaimer also warns that runtime inspection can change state and that generated conclusions may be wrong. Those are practical limitations, not footnotes to dismiss.

Why Morluto Cares About the Next Contributor

One of the quieter moments in the interview is Morluto's response to people making their first pull request. They say they try to acknowledge the work personally, because a first encounter with open source can determine whether someone comes back. The community did not arrive as a complete product strategy: Morluto says the Discord server was added only after the star count showed there were people to talk to.

That is a more durable interpretation of the viral rise than the raw count. A useful project creates a path from curiosity to a small contribution, then gives contributors enough feedback to improve it. Morluto does not pretend the tool is finished. They describe bugs and a large backlog, and imagine people adapting software for narrow, everyday frustrations such as games, printers and tax tools. Those are forecasts and examples, not demonstrated REA outcomes in this interview.

When Matt asks about anonymity and monetization, Morluto says they did not seek fame and would rather make room for other builders. They also mention ongoing academic research into agents and continual learning. The interview provides that account; it does not establish an announced REA business model or a released research result.

A First Responsible REA Project

Start small: one behavior in your own app that is hard to document, or one authorized artifact you want to understand before building an original improvement. Read the official README, use its linked installation guide, and review what setup proposes before changing your agent configuration. The README currently lists npx rea-agents setup as its quick start, but requirements and commands can change; do not trust a lookalike repo or copied install command. Try a bounded, low-risk static inspection before enabling runtime access.

The following prompt is an original business and product pilot, not a prompt from Morluto or a promise that REA can answer every question. It deliberately asks for permission and a distinct output rather than a one-to-one clone.

Interview Chapters

These are editorial navigation points drawn from the supplied transcript, not an official chapter list.

The person and the origin

What REA changes and where it stops

Open source and what comes next

Sources and Links

Checked 29 October 2026. The interview and supplied transcript are the source for Morluto's personal history, collaborators, motivations and forecasts. We have not reproduced their private experiments or independently verified the biographical details. The live repository documents the product as of this check; it may differ from the version discussed during recording.