Dragan Bajcic

Building cutting edge data/code/viz systems for 20+ years by day. Electronic music/digital art & ćevapi na kajmaku by night.

[Dragan Bajcic @kodisha]

Na repeat, 10000 puta.

Imogen ♥️

Young Marco 👑

[Dragan Bajcic @kodisha]

@pushmatrix @threadreaderapp unroll

[“paula” @paularambles]

if fable 5 agrees to do your work instead of downgrading you to opus, you’re in the permanent underclass

[Dragan Bajcic @kodisha]

@0xSero Lol wdym at launch, this shit came out not 24h ago 🤣🤣🤣

Already nerfed??

[kache @yacineMTB]

frontier coding abilities, as long as you're only working on react apps

[Dragan Bajcic @kodisha]

@OpenAIDevs Wait what, now Codex as well?!!?

https://t.co/OaQn54Hvqk

I built the same thing not 10 days ago - now cursor AND codex do the same thing in a same day!

[Dragan Bajcic @kodisha]

Wow @cursor_ai team built the same thing as I did - like 10 days ago!

I feel .... validated I guess 😎

https://t.co/LTYbHIz6gj

[Dragan Bajcic @kodisha]

@RealAngryDegen @ericzakariasson About ~48h for it to be useful and working.

Then I added few features here and there.

No users yet, I kinda build in open, just advertized it on few discord, asked for alpha users/testers.

You're welcome to join 🙌

[Dragan Bajcic @kodisha]

Wow @cursor_ai team built the same thing as I did - like 10 days ago!

I feel .... validated I guess 😎

https://t.co/LTYbHIz6gj

[Dragan Bajcic @kodisha]

@ericzakariasson Even though both sites have github like commit graph, I am at a slight advantage for having unlockable achievements 😎

[Dragan Bajcic @kodisha]

@ericzakariasson Even though both sites have github like commit graph, I am at a slight advantage for having unlockable achievements 😎

[Dragan Bajcic @kodisha]

@ericzakariasson Omg, Ii built a same thing, but for any/every agent (works with cursor as well)

https://t.co/mk8pE82XSc

[Dragan Bajcic @kodisha]

@rcanepag @argofowl @sama Double mog! 😎🔥🔥

[Dragan Bajcic @kodisha]

@slicknet I did, but somebody beat me to it :D

https://t.co/ME7VwqupSQ

☝️ 💯

[Dragan Bajcic @kodisha]

@SadClown idk, glupson.

codex + /goal ftw.

[Dragan Bajcic @kodisha]

@SadClown https://t.co/65ecsXNgX0

[Dragan Bajcic @kodisha]

@SadClown Hoće da ubaci co-authored by Anthropic Bot u commit

[Dragan Bajcic @kodisha]

@thdxr Not enough training data for Effect, they can't find chads for RL

[Dragan Bajcic @kodisha]

@rauchg Codex / Topchester
gpt-5.5
low - for imementation
xhigh - to write the plan

https://t.co/mk8pE82XSc

[Dragan Bajcic @kodisha]

Danas u Mladenovcu

[Dragan Bajcic @kodisha]

Bunch of people liked and bookmarked my post about "Planning Slices" skill, so here is the original file I use.

https://t.co/kRO4UCTWlC

You can also see here how I used clankers to optimize it:
https://t.co/mk8pE82XSc

[Dragan Bajcic @kodisha]

What works best for me is a good "planning slices" sklll.

My whole workflow for months now consists of:
- write an implementation plan
- imement plan (this became /goal recently)

The trick is that skill that writes the plan is optimized so that it tries to deliver a feasible end goal.

My skill doesn't just say "write a plan" but it instructs the agent to split the feature in smaller implementation slices, and then orders them in such order so that work is done from bottoms up, from contracts, types, validators, then up towards concrete implementation. Each slice has set of expected files to change, and what validation steps should be performed.

Then, and this bit helps the most, there is instruction to add any critical findings for the next slice to the plan doc it self.

Then as the goal runner starts the next slice, it gets all that and continues towards the end goal.

I have yet to run into a plan that cannot be fully implemented.

The workflow is:

- prompt to use planing slices skil <describe feature i want + LOTS of details> - ~ 3-5 minutes on 5.5 xhigh

- review, quick glance if nothing sticks out - 2min

- /goal implement docs/plans/xxx.md - 20-40min on 5..5 low/medium

😎🚀🚀

[Dragan Bajcic @kodisha]

@KingBootoshi @0xSero I see people are liking this, here is the skill it self:

https://t.co/kRO4UCTWlC

Enjoy, drop a star, or try ClankerLog 😎🥰

(see my profile)

[Dragan Bajcic @kodisha]

What works best for me is a good "planning slices" sklll.

My whole workflow for months now consists of:
- write an implementation plan
- imement plan (this became /goal recently)

The trick is that skill that writes the plan is optimized so that it tries to deliver a feasible end goal.

My skill doesn't just say "write a plan" but it instructs the agent to split the feature in smaller implementation slices, and then orders them in such order so that work is done from bottoms up, from contracts, types, validators, then up towards concrete implementation. Each slice has set of expected files to change, and what validation steps should be performed.

Then, and this bit helps the most, there is instruction to add any critical findings for the next slice to the plan doc it self.

Then as the goal runner starts the next slice, it gets all that and continues towards the end goal.

I have yet to run into a plan that cannot be fully implemented.

The workflow is:

- prompt to use planing slices skil <describe feature i want + LOTS of details> - ~ 3-5 minutes on 5.5 xhigh

- review, quick glance if nothing sticks out - 2min

- /goal implement docs/plans/xxx.md - 20-40min on 5..5 low/medium

😎🚀🚀