Hacker Newsnew | past | comments | ask | show | jobs | submit | 0123456789ABCDE's commentslogin

fwiw, response headers include: Last-Modified: Fri, 22 May 2026 19:08:09 GMT

yes, they do.

anyone who's, even slightly, clued into how agents access documentation, has been making changes to their pages. ex: https://searchtxt-web.fly.dev/search?q=aws


> I can definitively say llms.txt is not used by any AI players.

  https://developers.openai.com/llms.txt
  https://docs.anthropic.com/llms.txt
  https://geminicli.com/llms.txt
  https://github.com/llms.txt
  https://docs.aws.amazon.com/llms.txt
  https://openrouter.ai/docs/llms.txt

OP clearly meant that the AI players are not reading and/or honouring llms.txt of other websites when scraping.

i stand corrected, but what was clear to you, obviously was not clear to me.



Correct once again it seems.

beautiful visualization of "complex systems run in degraded mode"

https://how.complexsystems.fail/#5


What a great capsule of wisdom!

There is still a tendency within some parts of aviation (safety auditing) to look for root causes and use tools like "fish bone diagrams" despite the more holistic approach used after an actual crash or incident.


A bunch of different services on a single status page doesn’t make it a complex system. Most of these have no relation to each other other than the high level services on the cloud providers.

They're all part of the internet, which is one of the most complex systems ever built.

> A bunch of different services on a single status page doesn’t make it a complex system.

you're it does not.

> Most of these have no relation to each other other than the high level services on the cloud providers.

so, some of them are related to each other? some of them even share underlying infrastructure? perhaps multiple of these are considered infrastructure for some teams?

what is the point you're trying to make?


it wasn't stolen either. an employee accidentally included a source map file with the release.

why are folks looking at the output of the first pass?

my understanding, and experience, is that you 1. run a bunch of sessions with small permutations to create variety, 2. run more sessions dedupe reports into a smaller collections of potential vulns, 3. run a handful of agents at max effort to write PoCs + write-ups, 4. rank findings, 5. finally look at what, if anything that, was found. maybe ask questions, try and understand if the PoC is running against a realistic setup.

until you can confirm a vuln report is valid, you must assume it is invalid.


What Project Glasswing claimed at launch is that Mythos can "surpass all but the most skilled humans at finding and exploiting software vulnerabilities". What you're describing sounds more like making skilled humans more effective at penetration testing. That's cool, but it's not clear how much it matters, because most security teams were not previously bottlenecked on penetration testing capacity.

i wasn't thinking about pen-testing, but vulnerability-research, which seems to match that quote. but, you're right, gp is referring to "security scanning". i just feel like, even then whoever's running the research, should triage and validate results, before passing on to mgmt.

>Is the play to plan/design/architect with a heavier model than delegate structured tasks to these smaller ones?

always has been

claude code has opusplan — uses opus while in plan mode, switches to sonnet for execution.

https://code.claude.com/docs/en/model-config#opusplan-model-...

edit: you can make it work with sonnet for planning, and haiku for execution, or any other combination you fancy to work with.

https://code.claude.com/docs/en/model-config#control-the-mod...


seems to be a well understood angle inside anthropic — colossus use expands subscriber capacity; training goes to aws/gcp

https://x.ai/news/anthropic-compute-partnership https://www.anthropic.com/news/higher-limits-spacex


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: