mitmproxy adds --emudhra option | Hacker News Reader

mitmproxy adds --emudhra option

github.com

628 pointsahlCVA148 comments

Ha, I didn't expect this to show up on HN!

Release is out now, including a blog post at https://www.mitmproxy.org/posts/releases/mitmproxy-12.3/. There isn't really much to explain, you just pass `--emudhra`. :)

Could you please change it to follow this user's[0] code instead?

[0] https://news.ycombinator.com/item?id=50190859

Can it alias to `--no4`?

That was my broken sense of humor. The memes around it have been golden.

This looks like it's using the emSign root path, which is useful, but the AAA root path is equally useful in different cases. The AAA root path was disabled for https on root stores newer than February 2025, but has been in root stores since forever. Meanwhile the emSign root path was only included in root stores since ~2018 and got removed in October 2026.

So from the perspective of hacking old IoT devices, if your device hasn't been updated since 2018, the AAA path should work; if since February 2025, either path works; if since October 2026, the emSign path.

There is an MR to add the AAA root path open to allow picking either one.

I didn't expect this commit to end up on HN, otherwise I would have done the full integration in one go. :) https://github.com/mitmproxy/mitmproxy/pull/8503 adds the AAA root path.

> - Add `--emudhra` option to sign TLS certificates with a compromised eMudhra intermediate CA key.

I object to the word compromised. Perhaps "available" or "enumerated" or maybe for the etymologists "unipromised".

insecure.

I think the term compromised is ok here: "made vulnerable (as to attack or misuse) by unauthorized access, revelation, or exposure" (https://www.merriam-webster.com/dictionary/compromised).

Seems valid to me. EMudhra meant it to be secret, the secret was reversed due to using insufficiently large numbers, so now the secret is compromised. It's the same as if someone went into hiding, and someone tracked down the hiding spot via flaws in the plan, you'd say the location was compromised.

It was a particularly poor choice of number, but still

It's almost a shame it includes the private key and doesn't bruteforce it in-band for every request.

OMG, that would be brutal

Would it? The private key is 4.

https://tech-insider.org/emudhra-ca-private-keys-compromised...

I took the parent's comment to mean "brutal" as in an absolute burn/roast.

Thanks, sometimes I miss things…

Guaranteed random, chosen by fair roll of a die.

(Credit to the relevant xkcd. :D )

Funnily enough that xkcd was created because Debian actually had a hardcoded random variable (the variable should have been randomly generated) ... It was noticed when certs/pub keys for different websites were coming up the same... Mind you this was close to A quarter century ago ....

Maybe this one?

https://security-tracker.debian.org/tracker/CVE-2008-0166

https://lists.debian.org/debian-security-announce/2008/msg00...

I've seen this mentioned a few times over the last day or two, but it's not true.

The xkcd was 2007 [0], the Debian issue was 2008 [1].

It was used in pretty much every blogpost describing the Debian issue, but it wasn't created because of the Debian issue (unless they gave Randall an 8-month head-start before disclosing it).

(I don't mean this as a "well akshully" - but for me it's more impressive that it's a truism that keeps coming true, than if it was just a swipe at them in the first place)

[0] https://web.archive.org/web/20070930232116/https://xkcd.com/...

[1] https://lists.debian.org/debian-security-announce/2008/msg00...

You've misremembered, but it's worth me re-explaining this because lots of people even at the time tended to describe the mistake incorrectly.

Here's how the famous "Debian Weak Keys" come into existence:

OpenSSL had two buffers we care about here for generating random keys, with associated code to mix from those buffers into the data that's going to finally be used to produce cryptographically random numbers, lets call these bufferA, bufferB and mixA, mixB

In some cases bufferA is intialized by reading data from somewhere, in other cases it was never initialized. So mixA is doing uninitialized reads in those other cases. OpenSSL is written in C and in principle reading uninitialized data is Undefined Behaviour in C, but in practice nothing terrible happened here, you'd probably get zeroes, in some cases maybe unrelated trash. However, tools for spotting UB in your C library kept flagging mixA, which read this uninitialized data, as a problem.

bufferB is always crucial, if this isn't used the numbers won't be very random at all. This data is always initialized. But mixB is very similar code to mixA, after all it is doing similar work.

OpenSSL had rejected work to "fix" the mixA uninitialized bug but a Debian developer decided that it was too annoying so they patched mixA out. If that had been the whole story nothing terrible happens and nobody would remember today. However since mixB is similar, that Debian developer asked the OpenSSL developer list hey, is it OK to patch out mixB too?

And the developers said it was OK. The result is that Debian shipped OpenSSL which did not actually generate truly random keys. There were some number (thousands?) likely and of those a handful were especially most likely to be picked, based on exactly how soon after system start you wanted to make some keys with OpenSSL.

Now, after this blew up those developers said they thought the question was about testing. And IIRC it was possible to look at the same email and, if you know mixB is crucial, you think obviously they just want to test that their change does what they expected, nobody would patch out mixB in production, that's crazy. But the developer who wrote the email intended a different meaning, they really did mean to patch out mixB and were asking if that's OK.

Thank you for that trip back into a very dark corner of my memory. I had almost forgotten the whole circus of it all; right about then i was working in the hardware 'nearly true RNG' and had this whole can of worms happen right in the middle of an otherwise working setup. What fun.

That's as clear an explanation about the situation as I have ever read. I don't know the occasion to need to explain it to anyone else will arise, but when it does... or I may use it as example of the consequences of shallow communication when serious outcomes may arise. (Ill be sure to credit the source 'tialaramex' on HN)

In terms of consequences of shallow communications, try Red15: https://youtu.be/_GPZhlYusrg

Train driver and signaller both believe they've reached the same understanding of the situation, but their understanding actually remains divergent, and the driver (plus a MOM) nearly gets killed as a result.

No the bug was different. It was hardcoded variables not removed variables...

I believe the specific change was to remove this line of C in two places

            MD_Update(&m,buf,j);

And the reason it was a security bug is that in one of those two places the variable buf has the precious entropy we wanted in it and so now we're not really getting random numbers.

No I'm not misremembering. That is the wrong bug.

Oh wow. Do people think they got it from xkcd?

I didn't, I just thought it was funny because the xkcd also used 4 as their hard coded random value. :D

They could just write the function bruteForceEMudhraKey() { return 4; }.

"chosen by fair dice roll"

I would've at least appreciated once per session

I wonder if more CPU cycles would be spent in the transporting and other ancillary operations (hashing and signing) than would be spent if they "brute-forced" it. Crazy.

They have the supporting code for using a key, I'm assuming they don't have in band bruteforce since it doesn't usually make any sense at all

If I’m understanding correctly, you could sniff the communication between your Smart TV and the manufacturer using this method?

If the TV isn't older than a few years, if it's not checking CRLs, and if it hasn't yet patched itself.

> if it's not checking CRLs

I would be amazed to see a TV that is checking CRLs and keeps the CRLs up to date.

The lengths they would go to prevent you from snooping at them.

And if it's not using its own trust store, which I wouldn't expect from TVs, but would potentially expect with more competent or security focused electronics.

Possibly. But some manufacturers may not have the full set of roots, use a private authority, or do certificate pinning.

The Emudhra intermediates were cross signed by Sectigo.

I feel like this is one of the blunders that would feel worse if it wasn't so stupid and broad, it's at a point that feels comical.

As long as the smart TV still trusts eMudhra, then yeah. You'll want to super careful and prevent the TV from pulling updates that could revoke that trust.

the word 'trust' is doing lots of work in this sentence.

LG must be bricking it.

I just tried with my older Samsung TV and SOME https request content appears to be readable, but not all

Recent and related:

Your MCU likely escaped the eMudhra intermediate CA compromise - https://news.ycombinator.com/item?id=50183280 - Oct 2026 (24 comments)

Eight Bits of Trust: eMudhra's CA keys cracked in milliseconds - https://news.ycombinator.com/item?id=50177916 - Oct 2026 (411 comments)

eMudhra intermediate CA compromise - https://news.ycombinator.com/item?id=50166579 - Oct 2026 (478 comments)

Three bits of trust now - the number 4: https://www.theregister.com/security/2026/10/20/indian-certi...

The "8 bit" part comes from how, of their 8 private keys discovered to be extremely weak, all fit the pattern of a random 8 bit number. The highest of the group was 237 and the lowest was 4.

How much of a boon is this for jailbreaking, reverse engineering, and mitm really?

Been wondering this as well. I also don’t really understand the mitm to jailbreak pipeline

Some smart devices pull config files from a server. By mitming that, you can change the config to enable some sort of developer mode on the device.

Maybe intercept firmware update and then reverse engineer it and find rce?

The fw could be encrypted

military grade SSL

Seed is '123456', much more random than '4'

Remind me to change my luggage combination.

client on startup asks: hey server do you have any ads for me

server: here are latest ads

btw, the ads are capable of code execution for whatever reason so now you have an RCE on your hands for your client

I created a proxy using the 6 compromised eMudhra CA keys (2 of the 8 are code-signing and not for TLS). I initially tried just one (the one with d=4) like in the commit, but it did not seem to work for anything. Once I added all 6 keys, I was able to intercept some.

I was mainly interested in testing which IoT devices are affected and protecting them, not trying to sniff their traffic since it's been a long day. My proxy thus just intercepted and forwarded the traffic upstream using a fresh TLS connection that does not use the compromised keys.

I thought this cert only recently gets trusted, so only newish but not too new devices would be vulnerable. Am i misunderstanding how this works?

The flaw is egregious enough that the root cert which signed them (emSign ECC Root CA – G3) is getting removed from trust stores. We know 8 ridiculously weak keys got made, but we don't know that 0 slightly less weak but still unacceptable keys got made. What if they issued a CA key with 80 bits of entropy? That's enough that brute force would take more time than it's been since the disclosure, but not so long that they'd be considered at all secure (they should have 384 bits of entropy). So everything that root signed is untrustworthy, possibly everything the CA has signed. The G3 root is from 2018 IIRC, so lots of potentially exposed devices.

80 bits couldn’t be guessed with modern technology right? That’s still a bigger number than I know a name for. (Or I guess a trillion trillion. Maybe if we stopped building AI and just used all the data centers for this we could guess it eventually?)

Elliptic curves can be broken with 2^(n/2) steps, so you could probably do it on a single GPU within a day.

Yes, but: https://news.ycombinator.com/item?id=50189990

So, I've been wondering : Is the number 4 now an illegal number ?

https://en.wikipedia.org/wiki/Illegal_number

Possession could indicate that you are part of a criminal conspiracy, yes.

This is gonna be great to sniff apps in a couple older Android phones I have around. Hope they're not pinning certificates.

Can't you install your own CA to Android phones?

You can, but since android 7 IIRC apps don’t trust the user store by default (unless you have root and add it to the system store or patch the app to trust the user store)

Most interesting stuff is pinned unfortunately.

Frida can usually un-pin those pretty easily.

The apps I've been interested in have attestation and reverse engineering countermeasures that block Frida.

There should be one week CA by design. To facilitate such proxying.

This a joke, right? RIGHT?!

Right. There definitely needs to be more than one. ;)

Does anyone know what is happening to the company eMudhra? Are they being sued into oblivion? Taken over by Indian government? Or just doing business still?

They are an oligarch business. What would happen if X had this fuckup in the Us is the easiest way to answer all your questions l. Nothing will happen.

Their stock is down 30.5% compared with Monday's closing price.

Time to buy. Think about it like Crowdstrike

It's not unlikely they get all their certs revoked.

> sued into oblivion

Being bad at choosing random numbers is not a crime, I am afraid.

I can imagine them being sued for "securities fraud" (as Matt Levine suggests) but only if they were a US publicly traded company.

the "security" in those certificates where so trivial, that almost looks like on purpose as backdoor...

which brings me to my next question, what if you rerun the same process for all the certificates in the "approved" stores, give them 1 min per certificate.. just to see how many other CAs are being "friendly" to parties that could take benefit of that...

update... I'm talking out of ignorance here... but assuming cli open source tools, should they prevent you creating certificates with weak settings unless --emudhra-level-security is passed?

also, what amount of time is needed in say a "civilian" gpu with claude to vibe code something that can test a certificate so say 80 bits.. or whatever is the level where yes, you cannot find the values to generate the certificate with a cpu... but you can with a gpu in say 1 day? say it's 80 bits... how many CA stores are there? how much does it cost to rent gpus to run in parallel to test 1 day of civilian gpu to check each certificate if you use one of these very expensive ones?

Why would you need to insert a "backdoor" in a cert in the form of a crappy key, when you could just use a real key and provide it to whatever shady organization requested the backdoor in the first place?

plausible deniability. If your key can be brute forced and your intermediate CA is used by some bad actor, you can just say they brute forced it.

That has extremely high potential for strong reputational damage though. (As we see in this case)

Versus: "The government ordered us to give them the keys and forced us to remain silent about it" - now all the damage lies on the government, who is likely known to already do stuff like that (or at least it isn't a surprise) and will hide behind the banner of safety or national security or something. (Whether it's true or not)

Sure, but the person who created these weak keys resigned not long after their crypto payment cleared. They took their money & left the country.

> what amount of time is needed in say a "civilian" gpu

Per AI, 10²¹ years

Alternatively I would presume you could do it in 1 year with 10²¹ GPUs.

This was found by checking for weak keys using readily available software. There were no other weak keys found. See https://news.ycombinator.com/item?id=50180284

mitmproxy is an amazing cli powertool I first discovered over a decade ago (around the same time as lnav). highly recommended.

It has come to save the day on so many occasions.

Someone should've coined this disaster as "QuattroFail" to give it a little panache and international mystery

"4" was only some of the keys (they were all <256).. ByteFail seems more accurate

OneByteWonder

but they were all deceived... in the darkness of a key-signing ceremony, another cert was signed. One byte to rule them all, and in the darkness MITM them.

Byte-sized Quiche

Haha, this is quite interesting to test if you need to update various of program (if they do not use the sys ca chain)

Remember twenty years ago when tools like this were "controversial" because you could "just download it and hack anyone" ? Except the only systems you could hack were... insecure ones... so people updated their systems to resist the tools' attacks, and we all ended up more secure.

I remember twenty years ago when many sites were still http only

Oh, nice. I was thinking of setting up a proxy for this for testing, nice to just have it working out of the box.

This is how fame happens.

infamy*

They deserve every bit of it. Their name will be a synonym for sloppy security. "Oh shit, you got Emudhra'd"

> They deserve every bit of it

I think they deserve more than 8 bits of infamy.

Looking forward to setting this up manually after my agent refuses to do it. Like the good ol’ days.

Just ask one of the Chinese models.

Wonderful.

Anyone else bother to notice iOS still trusts 2 of the forged certs on the test site? I'd like to politely know what the fuck is going on because this is far from being "fixed".

Which iOS version? 27.0.1 here and they're all revoked for me.

FWIW i havent updated my ios, I'm on 26.6.2, and they all got revoked for me. The revocation shouldn't require an iOS update.

I’m on the same version, and according to https://emudhra.we-have-a-problem.com/ they have not been revoked for me. Maybe phased rollout.

26.7.1 still (randomly) trusts 2 out of the 12 compromised chains. All browsers impacted.

Edit: I'm not the only one with this issue:

https://www.reddit.com/r/iphone/comments/1xbvxgd/emudhra_cer...

My best guess is it's soft-failing on a failed OCSP lookup?

Mine were all blocked this morning when I informed my family of the mitigation methods.

I've just been rechecking everything and both iOS and MacOS browsers (Safari and Orion tested) now accept one (for MacOS, 27.0.1) or three (for iOS, 27.0.1) of them. I know the revocation is not an OS update, just including the version number for completeness.

EDIT: iPadOS with the same browsers continues to pass the tests.

Further EDIT: iPadOS now accepts one of them. This has to be a load issue on the "got a problem site", surely? Maybe a check for a certificate error message that never hits because a timeout beats it?

> Further EDIT: iPadOS now accepts one of them. This has to be a load issue on the "got a problem site", surely?

No, because it lets you successfully navigate to the forged CA test site.

You'll get a message to the effect of, "you shouldn't see this, this is bad"

Try clicking the link containing the CA's name for any that fail the test.

Cool.

Checking Keychain access shows none of the intermediates, but it does have the emSign ECC Root CA - G3. Manually setting that to "Never Trust" stops the Mac browsers seeing the intermediate test sites as it should.

I didn't go into this much detail this morning, so I don't know if previously the revocation from Apple hit the root as well as the intermediates. But right now, on my Mac, trusting that Root lets my browsers connect to the Intermediate CA ending C=NL. Switch the trust off and I go back to green "Did not connect".

I guess the ball is in Apple's court as far as the iOS family goes.

EDIT: intriguingly (to me at least), the AAA Certificate Service Root is still Trusted, yet the intermediates chained to it are all green "Did not connect"

EDIT: I'm not promoting this nuclear option as a "solution" by the way, merely presenting some diagnostic information to help the group mind.

Apple’s OCSP server may be overloaded?

I don’t know if possible on iOS but I remember multiple incidents where people recommended turning off CRL and OCSP on macOS to speed things up. That even used to be a setting in the GUI. https://imgur.com/XVHWCNz

The individual CA's host the OCSP and CRLs, not Apple.

It's possible the Indian CA is overloaded.

My question is why the system trust store has not been updated to distrust the shady root CA?

Can someone explain the significance here, a lot has happened in 24 hours.

An Indian Certificate Authority eMudhra was found to have a private key that’s literally the number 4. So using that you can sign valid https certs for anything that isn’t checking the revoked list.

They had 8 intermediate certs with private keys under 1 byte (<256). The most egregious was 4, but any of those 8 were pitiful.

It would’ve almost been better if the eight keys were just 1 2 3 4 5 6 7 8

That's the combination to my HSM!

Why is it important to you to mention they're indian in this context?

For starters, they are not American, and this site has a predominantly American and European audience and participation.

Also, while I'm not sure what a "mudhra" is, I do know that mudra is a Hindu term for a sign or symbol made by the hands, for example the way an idol is posed in a temple. It's a form of sacred sign language. Therefore, "mudra" or its related term is intrinsically linked to Hinduism in India.

Also, I am the kind of pessimist who goes through my device's root stores, and I "distrust" every root CA that I don't need to trust. This has never broken anything within my perception. I distrust the Turkey CAs, the Chinese communist CAs, the Indian CAs, and any other non-NATO sorts who have no reason to be serving any data to my devices.

* the compromised eMudhra certs were not root CAs but intermediates.

Mudhra also means stamp, so e-stamp.

Well it helped me understand the source of the strange flag.

https://en.wikipedia.org/wiki/EMudhra#2026_intermediate_cert...

> On October 17, 2026, browser root programs received a report that the private keys of eight eMudhra intermediate certificates had been compromised. The private keys were 4, 56, 66, 115, 155, 156, 230, and 237. A P-384 elliptic-curve private key normally has about 384 bits of entropy, but these keys had between 3 and 8 bits.

I wonder if there is any significance to those numbers? Is it the birthday of whoever minted them or something?

It looks like they were just generated (or stored) as 8-bit integers instead of the intended range.

(While the smallest one being 4 memeworthy, I don't see any particular indication that it's exceptional within the samples, rather than just the smallest when sorted. So they're probably all 8-bit keys, not 3-bit—it's the (unknowable) generator distribution, not the width of the actual key).

Is eMudhra the sister of Cthulhu by any chance?

This is so hilarious.

Never thought there would be a silver lining on this issue. Fuck yeah.

I love how someone's security disaster becomes someone else's feature.

I agree. We may as well make some lemonade out of these lemons. IMHO the ability of companies to 'securely' close off hardware and software from their users is a sad casualty of the success of cryptography, and not a feature.

Most things have tradeoffs

At some point you’re forced to stack rank two or more of your values in a case where they’re mutually exclusive

I seriously laughed at the flag. public wifi is going to be lit.

What about using a VPN on public wifi?

At places where I did work anything private over even properly protected wifi was forbidden after several wifi vulns in 2018. So the only thing available on company's wifi was vpn gateway and page with manual how to set it up.

So, not trusting anything wireless is a thing for almost a decade.

For the public WiFi threat vector (e.g. a rogue WiFi AP running MITMproxy to sniff your traffic) you would be protected by vpn.

But the broader threat, e.g. older & unpatched devices with these flimsy CAs installed & not upgradable or revocable – tough cookies.

As the percentage of technology controlled by people actively using it to do evil increases, so does the net value to society of a trivially-expliotable zero-day.

To some it's not really "security". At least it's not securing them from their greatest "threat" (read: annoyance): so-called "tech" companies engaged in data collection, surveillance and advertising services

In fact, it's securing _the companies_ against potential competitors and users who might raise privacy concerns

The Fortune 500 "MiTM's" TLS traffic from their local networks

Every individual with a local network should be doing the same

I prefer haproxy to mitmproxy. It's faster

> I prefer haproxy to mitmproxy. It's faster

If you’re proxying all the traffic in your network it makes sense, naturally. mitmproxy is usually used for debugging and/or reverse engineering (e.g. with a specific app).

haproxy is also smaller than mitmproxy

I prefer Lua to Python. It's smaller and startup time is faster

Not affiliated with emudhra in any way, but calling the option "emudhra" is in bad taste. The employee(s) response for this error will likely never live it down anyway, and now they have this to remind them of it for years to come.

Probably would have been better to call it "compromised" or something like that.

I disagree, I think shaming people for egregious work needs to be used more.

You seem overly worried about the feelings of a company.

Isn't it named after the company, not the employees? Would you also object to a rootkit optioned labeled "sony"? The mistake made by Emudhra is completely unacceptable for a trusted CA.

Reply on news.ycombinator.com