Plutkar

Practical AI for builders. Prompts, setups and before/after tests you can use today.

[Plutkar @Plutkar]

IT IS GETTING DANGEROUS: one sentence to a coding agent found 100 Dallas plumbers, checked their sites on a phone and built a ready pitch in about 18 minutes

how:
→ about a minute to shortlist 100 businesses in one trade and one city
→ 2 minutes to open the top 5 on a phone. 4 of 5 didn't repeat their own Google ad's promise, 3 didn't show their rating
→ 15 minutes to build a fixed page in their brand, a one-page proposal and the first email

the part worth stealing isn't the build. it's who you aim at: $2M to $20M companies, not one-truck shops. then one useful touch every 2 weeks to the best 100 for six months

in the case the article cites, nothing happened for 4 months. in month six, 28 bought at once

the full playbook with every prompt is in the article ↓

[The Boring Marketer @boringmarketer]

https://t.co/OViycR2WK9

[Plutkar @Plutkar]

@wherenox cars are where this gets risky though. a TV you can factory reset. brakes, airbags and your warranty you can't. the coding agent part is easy now, so liability is what automakers will lean on

[Plutkar @Plutkar]

@Frelzu launching it anonymously first was smart. 200T tokens of real usage before anyone knew the brand means the leaderboard spot wasn't bought with hype. curious how it holds up on long agent runs, that's where cheap models usually slip

[Plutkar @Plutkar]

@Qyravex the split is the right call. one tweak that helps: have the cheap model write a spec file with test cases, then give opus only the spec, not the original code. you get your own version instead of a line-by-line copy

[Plutkar @Plutkar]

@Ravlizk the rule that keeps this safe on every device: tell claude to disable apps instead of uninstalling, and to log every change with its undo command. one bad uninstall on a TV can mean a factory reset

[Plutkar @Plutkar]

@Hexflip the vizio numbers explain why TV makers won't let this slide. if ads pay for the hardware, expect firmware updates that bring the ad rows back. worth re-checking your setup after every update

[Plutkar @Plutkar]

@coolerwave the smart part is the billing layer. if every model call already runs through a payments company, usage-based pricing for AI apps gets way easier to ship. if the deal closes, my guess is that's the first thing they build

[Plutkar @Plutkar]

the prompt. put it in your claude project instructions:

"rules for every post you write for me:

1. never use: delve, foster, leverage, utilize, facilitate, empower, streamline, robust, cutting-edge, paradigm shift, tapestry, realm, beacon, multifaceted, meticulous, transformative, elevate, embark, supercharge, harness, ever-evolving, unlock
2. no em dashes. use periods, commas or a plain hyphen
3. no hashtags, no emoji
4. no "here's the thing", "the truth is", "what most people miss"
5. no rhetorical questions
6. first line under 12 words, with one specific number, name or result
7. no summary or one-liner at the end
8. cut any line that would still make sense in a post about a different topic

after the draft, list every rule you broke with the line number. then fix only those lines"

rule 8 does the most work. it deletes the filler

[Plutkar @Plutkar]

STOP USING THESE 22 WORDS. everyone can tell AI wrote your post

[Plutkar @Plutkar]

@elixlune "keeps the lessons for the next task" is the part most agent setups are missing. you can get a small version of this today: let the agent append to its own CLAUDE.md after every failed run, then review it once a week

[Plutkar @Plutkar]

@vayrenrose the research bottleneck point is the real one. a clean spec of how something behaves is worth more than the code itself. one thing worth adding: check the license and terms before shipping anything built from it, reverse engineering rules differ a lot by country

[Plutkar @Plutkar]

this is insane. claude code cleaned up a slow 2 GB android TV over wifi and free RAM went up about 90%

the part worth copying is the safety setup, not the speed:
→ it listed every app before touching anything
→ turned apps off instead of deleting them, so every change is one command to undo
→ uninstalled only 11 apps nobody used
→ never touched google play, the keyboard, bluetooth for the remote or the hdmi inputs

then it built a new home screen from scratch. big clock, their streaming apps, inputs, settings. about 12 MB of RAM

if your parents have an old android TV, this is a weekend project. the full prompt is in the article ↓

[Ansub @justansub]

My parents' old Android TV was slow and full of ads. Claude cleaned it up and built them a new home screen from scratch.

Here's exactly how, plus the prompt to do it on your own TV 👇 https://t.co/1ul2qKTmHe

[Plutkar @Plutkar]

one sentence in, a working interactive object out. free claude skill, MIT licensed

how to try it in claude code:
→ paste the install line from her page into terminal
→ restart claude code
→ type /colorfigures and describe an object

two prompts from her page worth starting with:
→ "a gumball machine where I can put in a coin and turn the crank"
→ "how does a toaster know when to pop? make it a Colorfigure"

the second one is the underrated use case. ask how something works and it labels every part. that's an explainer for a lesson, a product page or a support doc from one prompt

[April @Aprilzz423]

Many asked me how to make these, so I turned it into a skill~

Colorfigures: Describe an everyday object, and Claude draws it hand-inked and isometric, with buttons that actually work.

You can also ask how something works and the skill will label the parts to explain how each part does its job.

Free and open source 👇

[Plutkar @Plutkar]

@RetroRacing porting a 2024 demoscene sprite routine into a 1988 game used to mean a week of reading old assembly. did claude need the original comments or just the code

[Plutkar @Plutkar]

@patrick_kuhnke_ the difference is low-code locked you into a vendor. vibe coded apps are at least real code you can read, refactor or throw away. the maintenance problem is the same as any codebase nobody reviewed

[Plutkar @Plutkar]

@thdxr curious how the price/perf holds on long agent sessions vs short completions. that's usually where vendor pricing falls apart

[Plutkar @Plutkar]

@kimmonismus what lands with non-tech people is never the model name. it's watching their own boring task get done in two minutes. demos beat explanations every time

[Plutkar @Plutkar]

@wccftech oblivion remastered runs on unreal engine 5, and epic just shipped an official plugin that lets claude code drive the unreal editor directly. expect a lot more mods like this

[Plutkar @Plutkar]

BREAKING: claude code can now drive unreal engine directly. epic games shipped the official plugin

hundreds of editor tools across 30+ toolsets, and it's already in anthropic's official plugin marketplace

what it can touch:
→ actors, blueprints, materials
→ niagara vfx, control rigs, sequencer
→ state trees, umg widgets, gameplay ability system
→ c++ automation tests

the setup:
→ in claude code: /plugin install unreal-engine-skills-for-claude-code@claude-plugins-official
→ in unreal: enable ModelContextProtocol and AllToolsets
→ run ModelContextProtocol.StartServer in the editor console
→ first test: "list all actors in the current level"

prompts worth trying next:
→ "create a material instance from M_Base and make it 30% darker"
→ "run every automation test in this project and summarize the failures"

the part most posts skip, straight from epic's readme:
→ don't run it with --dangerously-skip-permissions. it can execute python inside the editor with no confirmation
→ commit before every long session. one tool call can move or delete tracked assets
→ localhost is not a trust boundary. keep the server off shared machines

on windows you need git bash or wsl for the session hook. the tools still work without it

repo link in the reply ↓

[Plutkar @Plutkar]

the prompt. put it in your claude project instructions:

"rules for every post you write for me:

1. never use: delve, foster, leverage, utilize, facilitate, empower, streamline, robust, cutting-edge, paradigm shift, tapestry, realm, beacon, multifaceted, meticulous, transformative, elevate, embark, supercharge, harness, ever-evolving, unlock
2. no em dashes. use periods, commas or a plain hyphen
3. no hashtags, no emoji
4. no "here's the thing", "the truth is", "what most people miss"
5. no rhetorical questions
6. first line under 12 words, with one specific number, name or result
7. no summary or one-liner at the end
8. cut any line that would still make sense in a post about a different topic

after the draft, list every rule you broke with the line number. then fix only those lines"

rule 8 does the most work. it deletes the filler

[Plutkar @Plutkar]

22 words that make people scroll past ai-written posts. this prompt bans all of them