gbrindisi

I also notice the growing trend to have EM carry individual contributor duties, I thought it was mostly a consequence of using coding agents but perhaps it's not: the EM figure as we know might just be a consequence of the golden zirp times (do you remember the endless technical EM vs non-technical EM debates?)

It's not just AI, these are the consequences of affiliate marketing. Just look at the crap that is YouTube nowadays.

I like the pattern of making a dedicated cli/harness and just build a skill to teach coding agents to use it.

At $work we built a thorough workflow to do security reviews, which is a pure skill to simplify adoption https://www.synthesia.io/post/automating-code-security-revie...

But the user experience is tricky because if we aim for very low false positives the run time for this kind of workflows is too long, it's then hard to justify blocking PRs.

I protest the modern web by trying to consume all content via RSS.

The feed reader shall be my main window to the world, and I am sorry that it's not obvious to content creators that I read them so I often send an email on the note of "I enjoyed this article you wrote, thanks".

I write a small blog myself and I see the other side of it, but I just gave up on SEO, metrics, etc. I want to be the change I'd like see in the world: I publish full content RSS, I remove all analytics, make the website as lean as I can, put out my contact data and my only success metric is # of interactions I get with occasional readers.

This is pretty much a spec driven workflow.

I do similar, but my favorite step is the first: /rubberduck to discuss the problem with the agent, who is instructed by the command to help me frame and validate it. Hands down the most impactful piece of my workflow, because it helps me achieve the right clarity and I can use it also for non coding tasks.

After which is the usual: write PRDs, specs, tasks and then build and then verify the output.

I started with one the spec frameworks and eventually simplify everything to the bone.

I do feel it’s working great but someday I fear a lot of this might still be too much productivity theater.

are agents/ still relevant after we got skills? I am genuinely confused on why I would need custom system prompts for specific agents, what should I use them for?

thanks for raising the alarm and sharing this, very insightful

(also beautifully presented!)

1. I dont have hard metrics at hand but with the latest Sonnet I'd say we reach consensus around 80% of the time, with Opus is almost always but we are not using it due to cost

2. The difference I see in agent behavior when they don't reach consensus is usually either

- when one of them didn't explore enough and lack context

- and/or when their risk assessment is off

The latest happen often, in other workflows based on agents we are now giving clear instruction on how to assess risk and where to draw a line to consider something a true positive.

3. validation is on Sonnet, we don't use persona based prompts but all the 3 validators get's the same task and context. The agent orchestrating them will take their output and make the final decision. We use an internal fork of the claude code github action for now.

I am doing something similar: I use openspec to create context and a sequential task list that I feed to ralph loops, so that i’m involved for the planning and the verification step but completely hands off the wheel during code generation.

I like openspec, it lets you tune the workflow to your liking and doesn’t get in the way.

I started with all the standard spec flow and as I got more confident and opinionated I simplified it to my liking.

I think the point of any spec driven framework is that you want to eventually own the workflow yourself, so that you can constraint code generation on your own terms.

fifteen years ago I use to do mobile pentests for banks and when we could not find anything significant for the reports we could’ve always count on “lack of rooting detection” and pin the risk on some vague mobile banking malware threat pushed by marketing. I am sorry I contributed to this nonsense.

100% security theater, and here we are.

ah I also did my own sandbox and at least twice the agent inside tried really hard to go around the firewall, so I ended up intercepting calls to `connect` to return a message that says "Connection refused by the sandbox, don't try to bypass".

Code here: https://github.com/gbrindisi/agentbox

the most annoying thing with Google Workspace is that you need super admin privilege to properly audit the environment programmatically, I believe because of the cloud-identity api.

I noticed that too and it’s kinda scary. Soon we will have the opposite of canceling, where the target will be deepfaked to say everything and its opposite to nullify their signal to noise ratio.

The crowdstrike incident taught us that no one is going to review any dependency whatsoever.

I wonder how far I could go with a barebone agent prompted to take advantage of this with Sonnet and the Bash tool only, so that it will always try to use the tool to only do `python -c …`

I can't use Google Meet on firefox/zen, I tried every setting combination I could find but the video call quality is still not comparable to chromium based browsers, so at work I reluctantly switched to Vivaldi.

If you figure this out please let me know!

We’ve kinda solved the detection of issues. what we still lack is understanding what’s important.

I think an underappreciated use case for LLMs is to contextualize security issues.

Rather than asking Claude to detect problems, I think it’s more useful to let it figure out the context around vulnerabilities and help triage them.

(for better or worse, I am knee-deep in this stuff)

This must be the best technical article I read on HN in months!

meta: if you use AI to write articles, don’t have them written so that I’m forced to use AI to summarize them

not in Italy

wouldn't this defeat the point? Claude Code already has access to the terminal, adding specific instruction in the context is enough

from the top of my head: various hacks to make apps available to spotlight, packages/apps behind their equivalents in brew to the point where I use nix to orchestrate brew for too many things, starting envs and build switch is too slow for my taste despite caching etc, nix the language is unfriendly and hard to debug, the stack traces are useless, etc

I do this too. Nix is incredible, until it isn’t and then I regret using it so much.

I’ll probably use something dumber for the next machine, and keep nix for servers and local vms.

I like it! How does your stack look like?

I run coredns with a blocklist, the config is like 4 lines.

What am I missing by not using AdGuard, PiHole and similar?

Current gen LLMs are trained on reddit, stackoverflow, etc. Discord is sitting on little data gold mine ready to be sold to train next gen LLMs.

the ultimate bikeshedding with dotfiles is home-manager with nix, an infinite cacophony of deep pain and enlightened pleasure

disagree, it is a good keyboard but the thumb cluster is not ergonomic. There is better, especially if you are up to build one yourself with a kit. Otherwise the voyager is already an improvement over the moonlander.

sometimes I stumble into stuff I’m sure I’ve already read something about in an article but if I didn’t bookmarked or made a note of it it’s very hard to find again (miniflux flushes content after a certain age)