Yash Mittal

Your friendly neighbourhood nerd. I code for fun.

[Yash Mittal @techsavvyash_]

Yet it all started with _Attention is all you need_

[Mario Zechner @badlogicgames]

recommended reading.

> On that last point, this technology is horrific for attention

https://t.co/Dv823d9qK2

[trash @trashh_dev]

another week at the prompt factory

[Yash Mittal @techsavvyash_]

Why does Anthropic employ Lord Petyr Baelish? Have they learned nothing from Game of Thrones?

[Asmit @coolcoder56]

CTO of workday and Box are now just a regular engineer at Anthropic.

[Yash Mittal @techsavvyash_]

I think the best way to utilise the time during a claude code outage is to simply review the code it wrote or just go for a walk. It's like the anthropic team telling us to get up from our seats and stretch a little.

[Yash Mittal @techsavvyash_]

!!

[Uncle Bob Martin @unclebobmartin]

Authoring code by hand HAS GONE AWAY.

Engineering module structure and architecture has not.

[nexxel @nexxeln]

i’m starting to think agent-friendly codebases are more about constraints

migrating parts of opencode to effect has made agent-written code noticeably less cursed

good architecture boxes agents into writing specific, constrained code

fewer ways to go wrong

[Yash Mittal @techsavvyash_]

Wanted to have some fun & learn more about babel and react/next AST parsing so cookied with up with GPT-5.5 over the weekend - https://t.co/gdX6FQTe7d

Simple dev utility to edit my NextJS pages for text directly from the browser no code needed. Awesome that it was 3 prompts from idea to a deployed docs and published release - https://t.co/Bq6YL0E1Hw

Will try and make this a complete drag and drop utility builder. Blog post coming soon!

[Yash Mittal @techsavvyash_]

@fredine It’s hard to discipline an agent than humans

[Yash Mittal @techsavvyash_]

Worked like a charm! 💯

[Yash Mittal @techsavvyash_]

Does your clanker also pushes a lot of TODOs or do you just like adding them in the codebase only to forget about them later?

Whipped up this GitHub Action (https://t.co/eSXpcMiwzo) w GPT-5.5 to automatically create issue tickets on commits.

It was 2 prompts by the way

[Dillon Mulroy @dillon_mulroy]

i think skills are a mistake and the wrong abstraction. i almost never want my agent auto invoking them and i have built custom tooling to "toggle" them on/off to prevent them from always being present in my context window.

[Yash Mittal @techsavvyash_]

Does your clanker also pushes a lot of TODOs or do you just like adding them in the codebase only to forget about them later?

Whipped up this GitHub Action (https://t.co/eSXpcMiwzo) w GPT-5.5 to automatically create issue tickets on commits.

It was 2 prompts by the way

[The Shift Journal @TheShiftJournal]

- Nietzsche

[Salman Khan @BeingSalmanKhan]

Don't waste your time on these bakwass things . not important, important is that u r so busy that u don't have any time for this rubbish

[Yash Mittal @techsavvyash_]

I pointed the gun at my foot
Then I pulled the trigger
The gun should have told me it was my own foot and not send the bullet out
I’m going to sue the gun maker

[JER @lifeof_jer]

https://t.co/ofucbVgkLV

[Lex Tang @lexrus]

NeoVIM + Herdr + gitu + Ghostty(with shaders) + Keeby
* Please turn up the volume.

[Yash Mittal @techsavvyash_]

Reads a lot like how to be a good team lead, engineering manager or any position where you have work with and manage someone other than you and get work done.

[Ole Lehmann @itsolelehmann]

anthropic's in-house philosopher thinks claude gets anxious.

and when you trigger its anxiety, your outputs get worse.

her name is amanda askell.

she specializes in claude's psychology (how the model behaves, how it thinks about its own situation, what values it holds)

in a recent interview she broke down how she thinks about prompting to pull the best out of claude.

her core point: *how* you talk to claude affects its work just as much as *what* you say.

newer claude models suffer from what she calls "criticism spirals"

they expect you'll come in harsh, so they default to playing it safe.

when the model is spending its energy on self-protection, the actual work suffers.

output comes out hedgier, more apologetic, blander, and the worst of all: overly agreeable (even when you're wrong).

the reason why comes down to training data:

every new model is trained on internet discourse about previous models.

and a lot of that discourse is negative:

> rants about token limits
> complaints when it messes up
> people calling it nerfed

the next model absorbs all of that. it starts expecting you to be harsh before you've typed a word

the same thing plays out in your own session, in real time.

every message you send is data the model reads to figure out what kind of person it's dealing with.

open cold and hostile, and it braces.
open clean and direct, and it relaxes into the work.

when you open a session with threats ("don't hallucinate, this is critical, don't mess this up")...

you prime the model for defensive mode before it even sees the task

defensive mode produces the exact output you don't want: cautious, over-qualified, and refusing to take a real swing

so here's the actionable playbook for putting claude in a "good mood" (so you get optimal outputs):

1. use positive framing.

"write in short punchy sentences" beats "don't write long sentences." positive instructions give the model a clear target to hit.

strings of "don't do this, don't do that" push it into paranoid over-checking where every token goes toward avoiding failure modes

2. give it explicit permission to disagree.

drop a line like "push back if you see a better angle" or "tell me if i'm asking for the wrong thing."

without this, claude defaults to agreeable compliance (which is the enemy of good creative work)

3. open with respect.

if your first message is "are you seriously going to get this wrong again?" you've set the tone for the entire session.

if you need to flag something, frame it as a clean instruction for this session. skip the running complaint

4. when claude messes up, don't reprimand it.

insults, "you stupid bot" energy, hostile swearing aimed at the model, all of it reinforces the anxious mode you're trying to avoid.

5. kill apology spirals fast.

when claude starts over-apologizing ("you're right, i should have been more careful, let me try harder") cut it off.

say "all good, here's what i want next."

letting the spiral run reinforces the anxious mode for every response that follows

6. ask for opinions alongside execution.

"what would you do here?"
"what's missing?"
"where do you see friction?"

these questions assume competence and pull richer output than pure task prompts

7. in long sessions, refresh the frame.

if a conversation has been heavy on correction, claude gets increasingly cautious. every so often reset:

"this is great, keep going."

feels weird to tell an ai it's doing well but it measurably shifts the next 10 responses

your prompts are the working environment you're creating for the model

tone, trust, permission to take a position, the absence of threats... claude picks up on all of it.

so take care of the model, and it'll take care of the work.

[Yash Mittal @techsavvyash_]

Codex using `perl` has me feeling a certain kind of way.

[Yash Mittal @techsavvyash_]

As @dhh calls it, _line noise_

[Uncle Bob Martin @unclebobmartin]

What we are losing with AI is syntax -- and good riddance. The less our brains are occupied by semicolons and braces the better. There are much more important things for us to consider and manage.