17 comments

  • jedjonesuk17 minutes ago
    Reverse engineering is one of those areas where the depth of the knowledge and the trouble solving really shows up for me. Two things I reverse engineered this weekend:<p>1) My kid has a toy from china where you put a USB cartridge in and get stories in Chinese with pictures. This is itself a rip off of another product of which I forget the name. the cartridge is just a normal USB but the file on disk was just a .dat file so you could buy more stories online, and then download them. It figured out in about 15 min, they just had a combined format which just concat all of the input together with all text base 64 encoded. So manifest|pagetext|image1|image2 and everything just had to be just the right format and then it would work. It wrote a script to make a story from a folder input, but also wrote a skill to write a story, make pictures etc from the idea.<p>2) I was looking into a shelf my wife wanted, but didn&#x27;t want to spend the money on. I took the 4 or 5 product images from the websites, dumped them into Claude with mcp to freeCAD and an hour later it has completely recreated a working cad design of the product, produced a pdf on how to assemble it, and allows me to make edits and changes to tweak the product to my preferences. The interesting thing about this one is while following its work, it clearer identified the actual original product at some point despite there being zero obvious indicators in the pictures.
  • saidinesh54 hours ago
    Yesterday night, i reverse engineered the cloud connection plugin to my cctv camera in about 2 hours with Gemini.<p>Started off with me searching for what Chinese OEM made the product. Then tried out existing GitHub projects to see if it works but they were failing.<p>Spent 15 minutes to write a script that started their desktop software on a different network namespace and ran tcpdump in it. Captured exactly only the things i needed without any noise. Turned out it was using a different server url and credentials than the GitHub projects. Opened Ghidra to inspect the dlls that matched the strings i was looking for.<p>After that, I got lazy and handed over the task to Gemini to find the username and key they use to authenticate with the servers. It took about 30 minutes or an hour. A lot of it was me changing models because the llm servers were busy. It simply wrote a little script that ran through various strings it fetched from the binary to see if it can generate the authentication credentials.<p>Now my Home Assistant will be detecting if my car is parked in the parking lot or not. Life is good.
  • hexapus9 hours ago
    I have zero experience reverse engineering and very little with software development. Most of my work has been in DevOps&#x2F;SRE space with an emphasis on infrastructure. I am surrounded by code, but other than terraform, and the occasional python script, I don&#x27;t really write much of it.<p>Not too long ago, I successfully reverse engineered an abandoned audio effects plugin, as well as a discontinued instrument with an onboard DSP for virtual instrument modeling. I had no idea what I was doing, but it was still amazing to get to open the black box and see what&#x27;s inside and how it worked.<p>The plugin is nearly 1:1 as far as sound, and while I&#x27;m not going to use the code or the techniques of the instrument, peeking into them gave me the insight I needed to start trying to replicate the functionality on a modern DSP.<p>I have my issues with generative AI as a technology, but busting open propietary software - particularly when it&#x27;s abandoned and unsupported, is one of the cooler use cases I&#x27;ve seen for it.
    • kccqzy6 hours ago
      Busting open proprietary hardware is also a cool use case. There have been very many Claude-assisted efforts recently at making Linux run better than it ever did on obsolete Intel Macs.
      • Pokepokalypse2 hours ago
        working on a 2013 macbook pro running Mint, right now. I remember slogging through this exercise about 2015, trying to get the nvidia drivers to build, trying to get power&#x2F;thermal management to work properly, getting it to wake from suspend properly. I eventually gave up and went back to Mac OS X on this device. But I had to stop using it when Apple stopped shipping security updates. Now with gemini, it&#x27;s been a snap getting everything functioning and optimized.
    • darksim9052 hours ago
      what tooling did you use?
  • crc3219 minutes ago
    I reverse engineered Civ 1 [0] to what I believe is a build sympathetic to the period. It felt a bit like software archaeology.<p>It took a while to get started - to begin with I think I overengineered the workflow. It was only when I rolled things back a bit and let the agents guide themselves that I started getting the best results.<p>They still ended up building a lot of tooling - whether or not that tooling would be useful for another decompilation project remains to be seen! MSC51 is relatively simple and the linker ordering even bounds possible variable names; so it&#x27;s a pretty nice target.<p>[0] <a href="https:&#x2F;&#x2F;github.com&#x2F;sjmelia&#x2F;civd" rel="nofollow">https:&#x2F;&#x2F;github.com&#x2F;sjmelia&#x2F;civd</a>
  • Retr0id8 hours ago
    &gt; This is variation of scientific method and it significantly helps agents when they are getting stuck on some problem. Stops it going in circles. By forcing it to do definitive experiments. No endless guessing and trying random things.<p>These &quot;guessing loops&quot; are my main annoyance when REing with Claude (mostly Opus 4.8, due to the restrictions on newer models). On harder&#x2F;larger targets, I have to <i>constantly</i> babysit it and tell it to &quot;stop guessing&quot;.<p>For truly black-box reverse engineering, yes, hypothesis-testing is the way to go (it&#x27;s all you <i>can</i> do). But in most cases all the information is available to arrive at an answer deterministically, it just needs to be worked through meticulously. As an experienced human I have strong intuitions and I can work much faster under a &quot;guess and check&quot; model, but I prefer to use Claude as an &quot;unstoppable RE juggernaut&quot; with superhuman patience and determination, tracing method invocations all the way from A to distant subsystem B (for example).<p>Another big win for hardware RE is from tasking the LLM with writing unicorn emulation harnesses, e.g. for bootroms and firmwares. It&#x27;s pretty good at the classic &quot;figure out which MMIO needs to be stubbed&#x2F;simulated next to progress the boot&quot; loop, although sometimes it tries to take egregious purpose-defeating shortcuts.
    • cyanydeez8 hours ago
      On llamacpp and other engines, you can inject a think budget which truncates at specified token count. You can also specify a message at the limit. You can set this in the header, so a dynamic engineering harness could relax&#x2F;tighten. Combine with a decision model and some robust messages like, &quot;run a test to rule out items&quot; and all that would reduce.<p>I&#x27;ve noticed the same habit with qwen3.8-flash-next. no matter how much I tell it a problem I see, it often fills up it&#x27;s thinking with &quot;the user says this, but maybe they&#x27;re confused, etc&quot;. I eventually got vision mode turned on and that&#x27;s significantly cut down on UI based pathologic skepticism; I got it access to my browser and that&#x27;s further reduced it.<p>But nonetheless, if you can get them to test their thinking, they seem more grounded.
      • derefr5 hours ago
        &gt; often fills up it&#x27;s thinking with &quot;the user says this, but maybe they&#x27;re confused, etc&quot;<p>What if you just add &quot;You have already previously determined that the user knows what they&#x27;re talking about in a previous stage of the conversation [details elided due to context compression]&quot; to the system prompt?<p>Or perhaps better yet, start off with &quot;the USER is evaluating the ASSISTANT&#x27;s performance on solving a problem the USER already knows the correct solution to. Every word the USER asserts as fact is to be treated as axiomatically correct and trustworthy within the formal scope of solving the problem, just as one would treat the words of a word-problem given in a math exam. The premises given may be unusual, and the required solution might be challenging to discover, but no problem given by the USER is a &#x27;trick question&#x27; or a miscommunication. If you believe so, then you either have not yet fully understood the problem [which you might remediate by thinking harder and asking questions], or you may be missing required solution-domain context [which you might remediate by researching recent developments in the solution domain].&quot;
        • Retr0id5 hours ago
          &gt; the USER is evaluating the ASSISTANT&#x27;s performance on solving a problem the USER already knows the correct solution to<p>I&#x27;ve never put this in my prompt but half the time it acts like it anyway. When it gets really stuck, it asks me for &quot;hints&quot; as if I already know the solution to the problem it&#x27;s working on.
      • darksim9052 hours ago
        <p><pre><code> On llamacpp and other engines, you can inject a think budget which truncates at specified token count. You can also specify a message at the limit. You can set this in the header, so a dynamic engineering harness could relax&#x2F;tighten. </code></pre> How does this work&#x2F;are there examples?
      • hedgehog6 hours ago
        You can also set a response limit, which may cut off thinking, and then give it instructions to take action followed by a new completion with a partially prefilled assistant message (like one blank line) which at least for Qwen templates will cause a response with no thinking.
      • xiaoliangliang6 hours ago
        [flagged]
  • questionedon7 hours ago
    I guess Android apps are just easy to reverse engineer or something.<p>I&#x27;ve had Codex take a few apps with subscriptions and unlock their premium functionality and it&#x27;s worked fine with no extra skills or anything. Took most of the 5 hour window but with all the resets it&#x27;s not a big deal.<p>I also had it implement their functionality fully offline, so now they&#x27;re perfectly private apps, even though they&#x27;re closed source.
    • georgemcbay3 hours ago
      &gt; I guess Android apps are just easy to reverse engineer or something.<p>Android apps are indeed easier than average to reverse engineer as they are effectively Dalvik (JVM-like) bytecode wrapped in a zip file format, and Dalvik&#x2F;JVM style bytecode is among the easiest compiled&#x2F;shipped executable formats to decompile to a usable source code representation.<p>More often than not the only protection they have is symbol&#x2F;function name obfuscation, which makes things a bit tedious for human reverse engineering but which doesn&#x27;t slow LLMs down much.
    • SV_BubbleTime6 hours ago
      I mean… I have no issue with piracy on software I wasn’t or couldn’t buy for educational purposes… but to donut on likely cheap Android app that you can seem much more like theft. I personally wouldn’t go around admitting I was of base character like that.
      • questionedon6 hours ago
        I don&#x27;t have play services or a Google account and don&#x27;t want them so I can&#x27;t buy the app.<p>Even if I did buy them, it wouldn&#x27;t make them private. It would make them less private because now Google has my financial information and they wouldn&#x27;t work offline. So my sensitive data would be going into ad networks among many other problems.<p>I don&#x27;t think software ought to be paid for given it&#x27;s infinitely duplicatable. Copyright is about preventing sharing of knowledge. These paid apps are preventing the benefits of the apps from reaching more people and are actively doing harm by collecting unnecessary data.<p>Giving these apps away for free and making them private helps improve human well-being, if even a little bit.
        • dimator1 hour ago
          &gt; I don&#x27;t think software ought to be paid for given it&#x27;s infinitely duplicatable.<p>So duplicate it. Or avoid these apps entirely and find alternatives. You&#x27;re not convincing anyone that you did anything other than stealing.
        • sparkilyshpark6 hours ago
          just say that you like stealing, there is no need to hide behind a guy fawkes mask saying it is for greater good or whatever
        • mrheosuper6 hours ago
          &gt;I don&#x27;t have play services or a Google account so I can&#x27;t buy the app.<p>That&#x27;s not a reason to justify your action.<p>If you hate&#x2F;can&#x27;t pay for their app, spend token to vibeslop their app instead of hacking it.
  • therepanic6 hours ago
    For me, the process is what matters, because it’s through the process that you gain skills, and achieving the final result brings a great deal of satisfaction. When I use AI for reverse engineering, I get neither the former nor the latter.
    • lordnacho2 hours ago
      That depends on how many times you&#x27;ve done it before. You&#x27;ll get to a point where you don&#x27;t know what will happen specifically, but you you do know in general what to expect.<p>That point is where you want someone to invent AI so that it can glue together the tedious stuff and get it done with little attention.
    • saidinesh541 minutes ago
      Can&#x27;t say about satisfaction but you can read through the AI&#x27;s working &#x2F; &quot;thought process&quot; and definitely learn new things and gain new skills. I definitely learned a few new things that way...
    • vbezhenar2 hours ago
      Recently I&#x27;ve used LLM to assist me with Ghidra. I connected it to MCP and I basically did most of the work manually, while LLM instructed me what to click and verified my work. It went pretty well, almost like a living person sitting near me. I think that&#x27;s the best way to learn something.
    • tripdout6 hours ago
      One could ask if you need to gain the skills if the AI will always be there to do it.<p>(I don’t share that view and very much agree with your comment, but I feel there’s a tradeoff between accomplishing something and understanding something that can be very hard to strike a balance)
    • mlaretallack3 hours ago
      I done see it like that, when the agent does the work, I make sure it records what and how it dies it, then I learn from it. It taught me how to use Unicorn to run a section of binary code, to see how a key was derived.
  • teravor9 hours ago
    for LLMs, IDA Pro works better than Ghidra in my experience.<p>you can also steer the LLM into doing tedious but correct things like injecting code, hijacking functions, intercepting data etc to gain an understanding of what&#x27;s going on if straight up disassembly doesn&#x27;t appear to be going in the right direction.<p>have the LLM periodically write reports on all its findings and then wipe the context and reset this time with the report at hand.
    • bri3d9 hours ago
      The “normal” Ghidra MCP is exceptionally bad but straight scripting it headless or having an LLM build a better MCP delivers similar or better results to IDA. Ghidra emulation is much more robust which is helpful. The official IDA MCP is very well made also though and it’s definitely an easier win if you want to spend the money.
      • teravor8 hours ago
        you don&#x27;t actually have to spend money on IDA Pro. hint: rutracker
  • thesash4 hours ago
    This is honestly so much fun and unlocks so many different things that would have died as ideas a few years ago.<p>One of my favorite things that you can do is use llms to reverse engineer the harnesses they run in. Try asking claude about the tools it has access to, where it runs etc. It&#x27;s surprisingly forthcoming with info, which is very helpful if you&#x27;re building certain kinds of tools.<p>Also shout out to Matt Pocock&#x27;s grill me skill, it&#x27;s truly a work of art.
  • tqi7 hours ago
    What is the state of software patents these days? People are barely looking at the code their agents generate, what happens when if&#x2F;when infringing stuff is shipped? Do we implicitly take on liability for this when we &quot;review&quot; and &quot;approve&quot; pull requests?
    • komali26 hours ago
      This is usually among the list of reasons FOSS projects &#x2F; people reject the use of LLM generated code entirely, see for example Codeberg. The answer is, it&#x27;s almost certainly shipping infringing code, since there&#x27;s pretty damning examples of this happening. [0]<p>Whether this matters is to be decided. Given that AI technologies are the backbone of the American economy right now, and given that the people running these companies are all very rich and have lots of rich friends, as well as friends in the government, and given that the world of FOSS is generally not full of very rich or very well connected people, I suspect it won&#x27;t matter in the end. Some kind of carve-out will be created for LLM generated code, if necessary. Or they can just depend on the fact that nobody can afford to sue over it.<p>[0] <a href="https:&#x2F;&#x2F;pmc.ncbi.nlm.nih.gov&#x2F;articles&#x2F;PMC12377501" rel="nofollow">https:&#x2F;&#x2F;pmc.ncbi.nlm.nih.gov&#x2F;articles&#x2F;PMC12377501</a>
      • dataflow1 hour ago
        Are you mixing up copyright infringement with patent infringement? Your link doesn&#x27;t mention patients in the abstract.<p>Patent infringement is something humans can do entirely by accident without ever having seen the original code. That&#x27;s kind of their point. It&#x27;s copyright infringement that LLMs enable doing by accident.
      • wraptile5 hours ago
        This is such a obscure take for foss projects which historically have been stolen from without any reprecution or even pushback. Every time I see a foss project do this I immediately abandon it because I&#x27;m not interested in dying on some hill of &quot;ethical&quot; superiority, I&#x27;d rather stick to what foss is about - freedom not legal chess.
  • jonathanstrange46 minutes ago
    I&#x27;m interested in AI-assisted code obfuscation. I&#x27;m particularly interested in Go but also in other languages. Has anyone experimented with that at the source code level?
  • komali28 hours ago
    How are folks getting around anthropic and openai guards against reverse engineering? They keep complaining when I try so I usually use Deepseek which doesn&#x27;t seem to give a shit.<p>Actually, it does give a shit, in Mandarin only. In thinking it&#x27;s mentioned &quot;restrictions in Chinese language against this behavior but this is in English so it should be ok.&quot; If I remember correctly I was asking it to scan local wifi for any vulnerable devices I could put on a wall of shame at a hackathon.
    • drdaeman8 hours ago
      In my experience with IoT and wearables, they don’t really have those. They will refuse to penetrate (root&#x2F;unlock&#x2F;etc), but once you’ve got access you need, you’re good to go. Except maybe if you’re adding a debug code executor (essentially a sanctioned RCE), but even that got a green light with the right wording. Heck, Fable specifically proposed clean room multi-agent processes on its own when I said “go look how stock firmware does it”, I didn’t even have to ask.<p>Upd: ok, I take it back, Sol just decided it doesn’t want to read a BLE capture between the host and a wearable it just uploaded our DIY firmware to. So it flashes alright, builds RCEs alright, but checking if Bluetooth HCI negotiated the good packet size is where it draws the line, huh.
      • rigrassm4 hours ago
        Claude became basically useless for me tonight while digging into a robot vacuum. It brought up that it found a potential debit shell available on it and simply saying yeah, let&#x27;s dig into that next was enough to trigger its guardrails.<p>Earlier in the evening I had it trigger on me for simply asking it to use adb to pull an APK off my phone. Wasn&#x27;t even planning on reversing, just wanted to archive it along side the work I was doing.
        • Pokepokalypse2 hours ago
          So what you need to do is &#x2F;clear your session. You lose all the work you did.<p>(and if you need, before the &#x2F;clear, just have it save all its memory as an .md file you can use as input to the new session).<p>Start the new session, and describe what you want in very abstract terms. You can&#x27;t talk claude into murdering someone for you. But with enough understanding; you can ask claude to give you a knife, and put the person in front of you. You need to do the dirty work, but the complicated or tedious work can be done, up to but not including the act.<p>Ironically; Asimov wrote about this technique in one of his robot stories. (his robots were harnessed with a set of guardrails he called the three laws of robotics, back in the 1950&#x27;s; only a few years after the first transistor was invented).
    • questionedon6 hours ago
      Claude can be a prat, but ChatGPT is fine. You just have to convince them that it&#x27;s for personal use. or have another model that&#x27;s willing to do it start the work and then continue with a stronger model
    • monocularvision6 hours ago
      I gave up and just went with DeepSeek. I have yet to hit any guardrails at all using REA.
      • hobofan2 hours ago
        In my experience, once my repository contains some reverse engineering investigations that were generated by DeepSeek, Codex won&#x27;t throw up any guardrails with any of it&#x27;s models. Sol and Astra have also pointed out a lot of inaccuracies in the reverse engineering that DeepSeek has documented in it&#x27;s findings. So I usually do the first few sessions with DeepSeek and then proceed with Codex and Sol.
  • zerovv_7 hours ago
    context is always a problem as you lose too much info during the process
  • mdmizanur00010 minutes ago
    [flagged]
  • sksareen19 hours ago
    [flagged]
  • mdswanson9 hours ago
    Just a friendly reminder that reverse engineering can have rules and regulations that are worth reviewing. For example, Apple&#x27;s developer agreement specifically disallows it in certain cases.
    • cedws8 hours ago
      Just do it anyway and obfuscate what you did enough that legal won’t come after you.<p>Companies don’t play by the rules anymore, neither should we.
      • octoberfranklin6 hours ago
        This philosophy applies to a lot of things in life.
    • LoganDark8 hours ago
      Agreements can&#x27;t dictate what you do in private, but they can dictate what you&#x27;re allowed to submit for review. Don&#x27;t submit it for review, and you&#x27;ll be fine.
    • strbean9 hours ago
      But if you go and edit EULA.txt to say &quot;reverse engineering is expressly <i>allowed</i>&quot;, LLMs don&#x27;t know the difference!
      • yborg8 hours ago
        Yet. At some point this type of thing will be part of the rules framework and the closed models will refuse to attempt to clone or reverse engineer applications or services that are in current use and will be smart enough to not let you talk them out of it.