Adam

engineering, open-source and other poor financial decision

[Adam @adamdotenv]

Everyone waits for automation to kill jobs.

What keeps you employed is the edge you create, not how much you produce.

If hiring one more person puts firm A ahead of an equal firm B, that job stays. However automated things get.

[Adam @adamdotenv]

@steipete if the openclaw 2.0 update + UI was done using Astra, AGI is not here yet

[Adam @adamdotenv]

@somewheresy I'm trying to do simple home automation for like 1 week already. Its unusable vibecoded slop, literally every single feature I've interacted with has obvious bugs, the UI being the worst one. The hype behind this is something i will never undertand.

[Adam @adamdotenv]

The innovations and ideas will always come from people who are deep in the problem space and actually doing things "hands on". Thats not implying you need to read code, but you need to deeply understand and be somewhat "cognitively involved".

Its hard to imagine people just staring at a blank wall and prompting an agent with brilliant ideas coming out of nowhere or just out of talking to an agent.

[Adam @adamdotenv]

@RhysSullivan crazy calling the entire continent of absolutely different nations an "european mindset" by actions of one company

[Guillermo Rauch @rauchg]

If you’re not reading the code, whether explicitly or through agentic inquiry, one or more of these is true:

○ You’re a beginner
○ Software is throwaway
○ You’re prototyping
○ You have no users / revenue
○ You’re taking on debt & risk
○ Your problems are basic

And btw. All of this is fine. But the reality is that models are still not at the “full autonomy” stage yet.

They make rookie mistakes, they go down bad architectural paths. I just had the best model in the world add a nonsensical 700ms delay to “settle” something and it told me “you’re right, I was cargo-culting” 🤨

I am on the camp that this need will diminish more and more. Most code is indeed going to be assembly-like. But we also have the global internet and software infrastructure riding on these models and narrative, and we have to respect that.

[Adam @adamdotenv]

The best results from review workflow I found so far:

Running 3 reviews in parallel using @nicopreme's pi-subagents

- thermo-nuclear-code-quality-review skill by @cursor_ai
- standards reviewer, from @mattpocockuk's code-review skill
- spec fidelity reviewer, same @mattpocockuk skill
- final synthesis step - weighted so simplification/readability wins rank first, sorted into "worth doing now" vs "overengineering" - a lot of review feedback is just forcing DRY needlessly

[Adam @adamdotenv]

@kushbhuwalka @superset_sh @conductor_build @vorfluxai @DevinAI @Cursor @herdrdev what about actual built in code editor so i can jump to func definitions and usage quickly? afaik only cursor does this

[Adam @adamdotenv]

@Ferm0s Agree, I recommend trying this plugin that combines herdr + nvim:
https://t.co/6uzE6w3OCj

[Kyle Mistele 🏴‍☠️ @0xblacklight]

the reason that nobody is using agents is because they are still wildly unrealiable even including the wildly expensive frontier models that are far too expensive to put in a mass-market consumer product

the reason it works for coding is because you have a six-figure-salary subject-matter-expert resource sitting there babysitting it all day (and still shipping slop)

[Adam @adamdotenv]

Creating your own agent orchestrator is the 2026 equivalent of baking sourdough during lockdown

[Adam @adamdotenv]

@carlassmann @rough__sea exactly, all the "ADEs" have such a shitty file viewers without LSP yet they rather release a mobile support then prioritizing this

[Adam @adamdotenv]

@rough__sea agentic orchestrators that have IDE inside - so i guess only cursor, zed and herdr + nvim

[Adam @adamdotenv]

@mattpocockuk It would be helpful to ship some complementary skill that will present the planned changes in some visual view, maybe inside /implement

[Adam @adamdotenv]

@minpeter This is very similar if you use herdr:
https://t.co/6uzE6w3OCj

[Panta @thepanta82]

The problem with "we're all managers now" is that no great piece of software was ever created by managers.

[Adam @adamdotenv]

unit tests are just garbage in 90% of cases

the only tests that matter in this age are high fidelity e2e tests

[Mitchell Hashimoto @mitchellh]

Btw this is why computer science matters. Yes, you can self teach it if you really want to (I personally wouldn't have had the discipline, so college was a way better force). However you learn it, the theoretical fundamentals matter and get applied to work every single day.

[Adam @adamdotenv]

@herdrdev Awesome, finally can see proper md or use plannotator right away

[Adam @adamdotenv]

@maxedapps I think this will possibly move into more visual HTML/Mermaid type of artifacts.
It seems way easier to just see once and then optionally dive deeper or "zoom in" into detailed view if needed

For example, I usually let the AI explain the possible paths like this:

[Adam @adamdotenv]

@draslan_eth @steida Agree, but currently its hard to do meaningfully complex work without looking atleast partially at the codebase and building understanding through pure reading of just agents output is exhausting.