🔍 Read the full analysis: Point Your AI Agent At Existing Software: A Guide To What’s Next on ThorstenMeyerAI.com
Get business pricing on tech for your team
- Business-only prices and quantity discounts
- Tax-exempt purchasing
- Multiple users, one account, clear invoices
TL;DR
A ThorstenMeyerAI.com guide published October 19 describes using the open-source Reverse Engineer Anything (REA) project to help coding agents inspect existing software and explain specific behaviors. Its central case study reports that a developer’s local accounting app matched QuickBooks Desktop sample-company statements account by account, but the result is a reported project outcome, not an independent assessment of REA.
Thorsten Meyer published a guide on October 19, 2026, explaining how developers can use the open-source Reverse Engineer Anything (REA) project with AI coding agents to inspect existing software and document how specific features work. The guide centers on a reported project that recreated accounting behavior from QuickBooks Desktop and matched financial statements for an official sample company, offering a practical example of using software analysis to inform a new application.
REA connects coding agents such as Claude Code, Codex, Cursor and Gemini CLI to analysis tools through a local Model Context Protocol server. It also includes a command-line utility and an agent skill intended to guide investigations. The project supports analysis of several kinds of software and files, including native applications, JavaScript and Electron apps, .NET and Android programs, firmware, databases and saved network traffic. For websites, it can observe activity in a browser controlled by the user.
Meyer’s central example is Ledger Reborn, a local accounting app built after investigating QuickBooks Desktop for Mac 2024. The case study says its profit-and-loss statement, balance sheet and trial balance for Intuit’s official sample company matched the original account by account, to the cent. The guide identifies investigation areas including inventory calculations, account-type effects on budget figures, reconciliation balances and cash-basis reporting. These are claims reported in the source article; it does not provide an independent audit of the project.
The guide recommends investigating one defined behavior at a time, then asking the agent to distinguish observed facts from inferences and unknowns. It also stresses that setup checks do not prove an agent is connected live: users should confirm that the agent can call an REA tool before trusting its explanations. REA version 6.7.0 was released October 18, according to Meyer.
Infographic 01 · Reverse Engineer Anything
Five places to point REA first, and how to tell when the answer is wrong.
REA gives coding agents such as Claude Code and Codex tools to inspect programs, apps and browser sessions. Each investigation needs a finishable question and a check.
npx rea-agents@latest setupThe five investigations
What to ask, and the check that catches a wrong answer.
- 1
Ask a question you can finish
One behaviour, one decision it informs. Example: how a partial payment changes an invoice balance.
It’s wrong ifThe agent can't walk through one concrete example with real inputs and outputs. - 2
Get REA connected and confirmed
Run the guided setup, then doctor, then call binary_session inside the agent before any real target.
It’s wrong ifAnswers arrive without a single REA tool call: you're reading the model's guesses. - 3
Turn a business rule into a spec
Inputs, stored fields, formula, edge cases, plus test cases checked against the original's own output.
It’s wrong ifYour version and the original disagree on a test case, even by one cent. - 4
See what has to survive a copy
Find what an app writes to the clipboard and what it keeps locally, then design your own copy and paste.
It’s wrong ifThe explanation doesn't predict what lands in a plain text editor, an email and your app. - 5
Your own app, your equipment, an old program
Start with software you own or rely on. Ask what it sends, stores or calculates when you do one thing.
It’s wrong ifIt only works for the example you watched. Test with a new file or a read-only command first.
Sources: REA on GitHub (stars via GitHub API, 19 Oct 2026); REA installation guide; REA security policy.
Powered by Thorsten Meyer AI
From Software Behavior to Specifications
The guide addresses a common gap in AI-assisted app development: an agent can imitate a program’s screens without reproducing the rules behind them. Existing software often encodes decisions about stored records, calculations and exceptions that are not visible in the interface. REA offers a way to gather evidence about those behaviors before designing a replacement or compatible tool.
That distinction matters for accounting and other rule-heavy software, where a plausible-looking result may still be wrong. A concrete walkthrough of inputs, outputs and conditions can help developers turn an observed behavior into a testable specification. It does not, by itself, establish that an analysis is complete or that a new application meets every legal, accounting or compatibility requirement.
As an affiliate, we earn on qualifying purchases.
How REA Fits the Investigation
REA is described as an MIT-licensed project on GitHub. Meyer reported about 325,000 stars and 82,000 forks on October 19, following attention on GitHub’s trending page earlier that month. Those figures are a snapshot from the article, not a measure of accuracy or adoption. The project’s maintainer is identified there as anonymous.
The article frames reverse engineering as a way to understand behavior and make implementation decisions, rather than simply copy an interface. It recommends questions narrow enough to answer with a complete example—for instance, how a partial invoice payment changes a balance and which account receives a posting. REA can help an agent inspect a target, but the developer still has to check the evidence and decide whether the explanation is sufficient.
There are also privacy and security limits. Analysis runs locally, but results are passed to the connected agent and may also be handled by its model provider under that provider’s data policy. The project warns that it is not a sandbox: opening an untrusted program exposes it to analysis tools running with the user’s permissions.
“This is not a sandbox.”
— REA project documentation, as quoted in Meyer’s guide
software reverse engineering software
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Limits of the Reported Results
The source article presents the QuickBooks comparison as a project case study, but the material provided does not include an independent review, reproducible test results or a full account of the comparison method. The reported match to the cent applies to the statements for the specified sample company; it does not establish accuracy across other companies, transactions, QuickBooks versions or edge cases.
It is also unclear how reliably REA performs across its broad list of supported software types, what practical limits apply to each analysis engine, or how often agent-generated explanations need correction. Meyer notes that claims should be marked as observed, inferred or unknown, but the guide does not establish that following this process guarantees a correct result. Model-provider data handling depends on the provider’s own policy.
accounting software analysis tools
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Verify Setup Before Analysis
Meyer’s immediate recommendation is to check prerequisites, install the agent connection and verify that the coding agent can call an REA tool before opening a target. The guide says REA requires Node.js 22.19 or newer; native-program analysis may also require a disassembler such as Ghidra, Hopper or IDA. Its setup utility can check files and prerequisites, but users are advised to confirm the live connection separately.
After setup, the suggested next step is to test analysis on a harmless application the user owns, then choose a narrowly framed behavior and ask for a complete example with its evidence and uncertainties identified. The article gives no announced independent evaluation or future release date. Further evidence about the QuickBooks case, REA’s accuracy on other targets and its handling of different software types remains to be established.
As an affiliate, we earn on qualifying purchases.
Key Questions
What is REA?
Reverse Engineer Anything is an open-source project that connects coding agents to software analysis tools through a local MCP server, alongside a command-line tool and an agent skill.
What did the QuickBooks case study report?
The guide says Ledger Reborn matched QuickBooks Desktop for Mac 2024’s profit-and-loss statement, balance sheet and trial balance for Intuit’s official sample company, account by account and to the cent. The supplied article does not include independent verification of that result.
Does REA keep all analysis data local?
REA runs analysis on the user’s machine, but its results are passed to the connected AI agent and may be handled by the model provider under that provider’s data policy.
How can users tell whether the agent is connected?
The guide says the setup diagnostic checks files and prerequisites, not the live agent connection. Users should confirm inside their agent that it can call an REA tool before relying on its answers.
Is REA safe for unknown programs?
The project warns that REA is not a sandbox. Meyer’s guide advises treating unknown binaries like other untrusted files because analysis tools run with the user’s permissions.
Source: ThorstenMeyerAI.com
Halloween Picks
halloween
As an affiliate, we earn on qualifying purchases.
