HNHacker News
TopNewBestAskShowJobs

alexchantavy

1,023 karma · joined March 7, 2015

I work on a tool called https://cartography.dev and am building a company around it called https://subimage.io
submissionscomments
alexchantavy··on Show HN: Doom port to pure Go – Gore
I'd love to `kubectl run doom` and kill monsters when I'm supposed to be working
alexchantavy··on GitHub is no longer independent at Microsoft after CEO resignation
It's always better when companies are hungry for business. I thought that in 2016ish it was super cool for Microsoft to get into Linux, build VS Code, and make bets like the Surface Studio.

For comparison, I think Mac OS in 2008 was also at a bit of a golden age:

- You had native file support for .iso, .zip without needing to install crapware like Winzip.

- You even could preview *.psd files out the box.

- You had first-party apps like Image Capture to scan documents without needing to install extra software.

- There was an amazing third-party app ecosystem with things like Yojimbo, OnyX, Little Snitch, Quicksilver, Handbrake, Coda, Adium.

This was around the time of the "I'm a Mac" campaign when Apple was _hungry_ to win business away from Windows. All of these small, polished advantages made me fall in love with the experience.

OSX today is still good but there definitely isn't that same level of "underdog hunger" showing up in the products as of late.

Anyway I'm just trying to say companies being hungry for business shows up in its products and that's better for consumers.

alexchantavy··on Things that helped me get out of the AI 10x engineer imposter syndrome
I didn't say that. When you're trying to get a job done, it's time consuming to sift through a long tutorial online because a big part of that time is spent determining whether its garbage and whether its solving the exact problem that you need to solve. IME the LLM helps with both of those problems.
alexchantavy··on Ask HN: Have you ever regretted open-sourcing something?
If this is something that you want to solve, I've found that having a PR template checklist that asks contributors to include tests helps a lot. Say something to the effect of

[ ] Added unit tests

[ ] Included a screenshot of the code working (this helps reviews go faster)

alexchantavy··on Things that helped me get out of the AI 10x engineer imposter syndrome
And nowadays if you type that into a search engine you may be overwhelmed with ads or articles of varying quality that you'll need to read and deeply understand to adapt to your use-case.
alexchantavy··on Air Force unit suspends use of Sig Sauer pistol after shooting death of airman
If there is a topic, you will find someone on HN to go deep on it. I love it
alexchantavy··on Don't animate height
Is there a good way to learn techniques like this from first principles other than trying to become an employed frontend dev?
alexchantavy··on Malware found in official gravityforms plugin indicating supply chain breach
Yeah that's what's called an assume breach/zero trust mindset. In a modern environment you can't rely on the network perimeter being a security boundary, so you need to minimize permissions (so that if an identity is hacked then the blast radius is reduced) and invest in detections and remediation plans.
alexchantavy··on Why does my ripped CD have messed up track names? And why is one track missing?
Wait so back in the day I remember Winamp let you configure a CDDB thing and it connected to something called.. Gracenote? (Am I remembering that correctly?) iTunes desktop at some point used to handle this all for you and I assumed it was pulling from those sources under the hood. Where did MusicBrainz come from?
alexchantavy··on Why Writing by Hand Is Better for Memory and Learning
+1, most of the time in college I was so upset how professors would just have a million words and equations on a slide and expect us to look at it later.

I learn so much more when a professor has a whiteboard and writes out an equation is derived one step at a time.

alexchantavy··on Trying to teach in the age of the AI homework machine
> I'm 62, and what allowed me to follow my curiosity as a kid was that the school lessons were finite, and easy enough that I could finish them early, leaving me time to do things like play music, read, and learn electronics.

I see both sides of this. When I was a teenager, I went to a pretty bad middle school where there were fights everyday, and I wasn’t learning anything from the easy homework. On the upside, I had tons of free time to teach myself how to make websites and get into all kinds of trouble botting my favorite online games.

My learning always hit a wall though because I wasn’t able to learn programming on my own. I eventually asked my parents to send me to a school that had a lot more structure (and a lot more homework), and then I properly learned math and logic and programming from first principles. The upside: I could code. The downside: there was no free time to apply this knowledge to anything fun

alexchantavy··on Build real-time knowledge graph for documents with LLM
I built an open source one (https://github.com/cartography-cncf/cartography) and am building commercial support around it (https://subimage.io)
alexchantavy··on Build real-time knowledge graph for documents with LLM
IMO knowledge graphs are a must have for security use-cases because of how well they handle many-to-many relationships. Who has access to read each storage bucket? Via which IAM policies? Who owns each bucket? What is the shortest possible role-assumption path available from internet-exposed compute instances to read this bucket? What is the effective blast radius from a vulnerability that allows remote code execution on an internet exposed compute instance?

Or, I have a docker container image that is built from multiple base images owned by different teams in my organization. Who is responsible for fixing security vulnerabilities introduced by each layer?

We really could model these as tables but getting into all those joins makes things so cumbersome. Plus visualizing these things in a graph map is very compelling for presentation and persuading stakeholders to make security decisions.

alexchantavy··on PyPI Organizations (2023)
PyPI is such an important service and as a Python user it's easy to take for granted that it just works. I recently had to make a config update from my project's GitHub repo to PyPI and lost the password and had to do account recovery, and then suddenly realized "wow, they take care of a lot of other orgs", and "wow, this is a TON of ops work" -- see the issues _just_ on account recovery: https://github.com/pypi/support/issues.
alexchantavy··on OpenAI reaches agreement to buy Windsurf for $3B
100%. You're referring to Cursor?
alexchantavy··on OpenAI reaches agreement to buy Windsurf for $3B
Yeah, in the recent Lightcone Podcast episode, Varun was talking about how they have a lean eng team but large sales org. I thought that was super interesting for a dev tool since I was expecting a dev tool to involve bottom-up sales to the dev instead of top-down sales to a leader like a CTO or VP of Eng
alexchantavy··on Will the Humanities Survive Artificial Intelligence?
I actually think humanities become more relevant than before in tech with AI. For example, good prompts for image generation remind me of authors setting the scene for their novels, so being trained to do this well is an advantage.
alexchantavy··on Launch HN: Cua (YC X25) – Open-Source Docker Container for Computer-Use Agents
There are companies like https://www.keycard.sh/ taking this on. There are other competitors too but I can't think of them atm
alexchantavy··on Launch HN: mrge.io (YC X25) – Cursor for code review
Been using this for https://github.com/cartography-cncf/cartography and am very happy, thanks for building this.

Automated review tools like this are especially important for an open source project because you have to maintain a quality bar to keep yourself sane but if you're too picky then no one from the community will want to contribute. AI tools are like linters and have no feelings, so they will give the feedback that you as a reviewer may have been hesitant to give, and that's awesome.

Oh, and on the product itself, I think it's super cool that it comes up with rules on its own to check for based on conventions and patterns that you've enforced over time. E.g. we use it to make sure that all function calls that pull from an upstream API are decorated with our standard error handler.

alexchantavy··on Claude Plays Pokémon
Haha yeah this is cool but the days of watching Twitch Plays Pokemon or RNG Plays Pokemon or things like that were much more entertaining
alexchantavy··on Launch HN: SubImage (YC W25) – See your infra from an attacker's perspective
We're looking for customers who find value in Cartography but don't have the resources to self-host. Open source is big in helping us meet people and learn the needs of teams who would be interested in commercial support. For pricing, it varies because we can take on very different infra requirements depending on the size of the customer's environment or their data freshness needs.
alexchantavy··on Launch HN: SubImage (YC W25) – See your infra from an attacker's perspective
> I understand that you are very early in boot-strapping, but what I was missing while skimming over the videos and links and webpage is a better high-bird view or contextualization of the apporach.

For a higher level view and contextualization, can you share more on what you mean? This would help give us a better idea on what to build.

> I was considering a demo, but the two options (chat and quick chat) were a bit unclear to me what they would archive / how they are structured.

Ah, you're referring to our cal link (https://cal.com/team/subimage)? It's basically up to you -- we can show you something in 15 minutes or 30 minutes based on your availability and based on what you're interested in -- would love to hear feedback in call!

alexchantavy··on Launch HN: SubImage (YC W25) – See your infra from an attacker's perspective
Thank you! Our business model is B2B software as a service. We're offering a fully-hosted offering around Cartography where we add useful features that enterprises want like automatic fix actions, recommendations, a natural language interface, and others.
alexchantavy··on Launch HN: SubImage (YC W25) – See your infra from an attacker's perspective
Not at all, super appreciate the discussion and your detailed read!

> what I'm asking is whether you envision your future as one of ("eh, it's good enough", "we're integrating more libraries that attempt to formalize IAM", or "we'll roll our own policy engine in python, how hard could it be")

It's a combination of 2 and 3. Today we use the policyuniverse library for things like s3 bucket-policies, and we have that self-rolled policy engine described in that blog post I shared. I should've mentioned earlier that this is the first time I've seen IAMSpy, thanks for sharing.

I think we're currently pretty good at permissions evaluation since that feature gives lots of value as it is, but there is a lot more to do. Continuing to improve this and being able to connect that with other data is a priority since it's one of our main value propositions.

> there is a massive opportunity for expectations being out of alignment unless you have a plan to get from where you are to Industrial Grade Introspection.

Industrial Grade Introspection is absolutely the plan. I'll also add that we're especially interested in highlighting cases that involve permissions that go between providers - like Okta->AWS, Opal->AWS, etc - and we intend to be very competitive here.

alexchantavy··on Launch HN: SubImage (YC W25) – See your infra from an attacker's perspective
Makes sense -- we're focused on fixing problems over just being yet another Jira ticket generator.

> Found Splunk creds in a log? Awesome, start using them. Syslog in an S3 bucket... boom. You are now hitting the stuff that every other ASM/visualization tool has missed.

This is my dream :). This past weekend I was playing around with something where if I clicked on a SecretsManagerSecret node then it'd give me the CLI commands to assume the roles and then retrieve the secret. It'd be neat to take it a step further and be able to click here and get a shell -- I don't think we're _that_ far off from that (but for now to be very clear we're focusing on read-only actions only since a security tool with permissions to do scary things in your environment kinda defeats the purpose).

alexchantavy··on Launch HN: SubImage (YC W25) – See your infra from an attacker's perspective
> have more insight into what, actually, the IAM Roles could do

We 100% do this, see https://eng.lyft.com/iam-whatever-you-say-iam-febce59d1e3b.

We evaluate the policies for the IAM principal against the resources to determine what actions they can perform on each resource. This is configurable too; here's the set of the default permission relationships shipped in OSS: https://github.com/cartography-cncf/cartography/blob/master/...

It doesn't cover conditions since those can be wacky complicated, and it doesn't cover resource policies (yet!) but in my experience this is still a very good heuristic that is already more accurate than AWS IAM Analyzer when I played with it.

The next step we're working on is to take this access map and correlate it with event data to see which permissions are used/unused so that we can prune them for ensuring least privilege. More to come here.

Edit: adding on for the part of your question about what features are paid or OSS, our paid offering is fully hosted and includes things like automatic suggested fixes, a natural language interface, customization with our dynamic schemas, and other bells and whistles. I'm not a fan of doing things like premium modules because I don't want to ever get in the position where I'm declining a pull request in open source because it covers a premium feature; that doesn't feel right.

alexchantavy··on Launch HN: SubImage (YC W25) – See your infra from an attacker's perspective
AWS has Config to give you an inventory, but that only covers AWS. My guess is that there's not much incentive for the major cloud providers to build a product to help you correlate data across other products.
alexchantavy··on Launch HN: SubImage (YC W25) – See your infra from an attacker's perspective
Depends on how small. Most seed stage and early companies are more worried about product-market fit and not security. For now, we're probably best fit for companies building out their first security teams, so that's _usually_ series A and later. That said, there might be something there, I'm open to figuring something out for a smaller company.
alexchantavy··on Launch HN: SubImage (YC W25) – See your infra from an attacker's perspective
Thank you, this is very helpful especially given your experience in the space. I intended to frame this like "there are many tools that let a security team can pull in data from the cloud providers and detect misconfigurations, but this becomes soo much more useful when they're able to contextualize it against their internal data". If I'm responding to log4j, I want to know all of the services that are running that affected library, which ones are internet open, and who in the organization owns it. That last part is key for actionability.
alexchantavy··on Launch HN: SubImage (YC W25) – See your infra from an attacker's perspective
> does SubImage actually recommend changes and generate terraform?

We recommend changes, though we don't generate Terraform just yet. Great feedback on the specific fixes for this case, thanks.

← PreviousPage 4 of 12Next →