Sabitlenmiş Tweet
Marchel
4.8K posts
Marchel
@marchelfah
Founder @wazzapihq | Malang - Bandung 👋 | Sometime English - kadang Bahasa Indonesia
Bandung, Jawa Barat Katılım Ocak 2019
841 Takip Edilen938 Takipçiler
just came across kronos. a foundation model trained on financial market data. feels like the right direction for stuff that general models dont handle well github.com/shiyu-coder/Kr…
English
wazzapi mcp server is live. my ai agents send wa messages from claude code or cursor. no manual api calls. just describe what to send. docs.wazzapi.com
English
14x yoy commit growth is the real signal here. ai made building cheaper and now more teams build. more code always means more people needed to own it14x yoy commit growth. ai made building cheaper so more teams build. more code needs more people to own it
David Sacks@DavidSacks
Q: How are job postings for software engineers rising rapidly despite AI agents automating coding? A: Because there’s far more code to manage than ever before. We’re already seeing a 14x YoY increase in GitHub commits, and it’s accelerating. AI has dramatically lowered the cost of writing code, so it’s now being used across far more businesses, applications, and use cases. We’re at the beginning of a massive productivity boom driven by the proliferation of bespoke software throughout the entire economy. Coding has been AI’s breakout use case this year. The fact that it’s increased demand for software engineers — rather than decreased it — should call into question the entire “AI will cause mass job loss” narrative.
English
every wa api makes you sit through a sales call. wazzapi you just sign up, pip install, and send. built by someone who actually codes. wazzapi.com
English
It's gonna be really funny when Mythos drops and is roughly on par with GPT-5.5 at 10x the price
English
you can pip install heretic-llm and run it on any instruct model yourself. the uncensored model space is moving fast. github.com/p-e-w/heretic
English
hit this wall a lot on bigger projects. agents lose context and start guessing. superpowers adds structure where agents need it most. 209k stars. github.com/obra/superpowe…
marcus@marcusyul
Acaban de convertir Claude Code, Codex y Cursor en un senior developer real. Gratis. Se llama Superpowers y ya tiene 204.000 estrellas en GitHub. El problema con los agentes a partir de cierto tamaño de proyecto es siempre el mismo: Van directo al código. Debuggean a ciegas. Abandonan los tests. Y cuando el proyecto crece, empiezan a alucinar. Superpowers lo resuelve forzando el proceso de un ingeniero senior: → Brainstorming siempre antes de escribir una sola línea → TDD red/green sin excepciones (borra el código si no hay test previo) → Subagentes en paralelo con git worktrees aislados → Contexto bajo control durante horas de trabajo autónomo El flujo ya no es "prompt → código → rezar": Brainstorm → Spec → Plan → TDD → Subagentes → Review → Ship Cero código antes de tener tests pasando. Compatible con Claude Code, Codex, Cursor, Gemini CLI, OpenCode y Copilot CLI. 100% open source. Licencia MIT. Ya no escribes prompts a ciegas. Tu agente trabaja como un senior. Os dejo el link de la repo aquí ⬇️
English
people ask why wazzapi uses flat pricing. per message fees get more expensive the more you send. thats broken. wazzapi.com
English
people ask why wazzapi uses flat pricing. charging per message means charging more as you grow. that punishes success. wazzapi.compeople ask why wazzapi uses flat pricing. charging per message means charging more as you grow. that punishes success. wazzapi.compeople ask why wazzapi uses flat pricing. charging per message means charging more as you grow. that punishes success. wazzapi.comadded auto warmer, campaign manager, and mcp support to wazzapi. the wa api i wished existed when i was building. wazzapi.com
English
@ClaudeDevs security feedback right in the editor while you type. catching vulns during dev beats finding them after a build scan
English
We’ve shipped a security-guidance plugin for Claude Code that helps identify and fix vulnerabilities as you’re writing code.
Available for all Claude Code users. Install from the plugin marketplace (/plugins).
English
theo spent 2 months figuring out gpt-5.5 instead of giving up after one bad session. thats the actual skill most people skiptheo spent 2 months figuring out gpt-5.5 instead of giving up after one bad session. thats the skill most people skip
Theo - t3.gg@theo
It took me like 2 months, but I've grown to love gpt-5.5. You have to prompt entirely different and put some time into your agents[.]md. Now that I'm over the hump, I can't really use any other model for code.
English
founded wazzapi to fix what i hated about wa apis. flat pricing, sdks that work, mcp support, no sales process. still the same mission. wazzapi.com
English
@_itsjustshubh yeah exactly. once you stop anchoring to one specific tool the whole agent space opens up. same config works across all of them and that freedom is underrated
English
@marchelfah portability is genuinely underrated. spent too long locked into tool-specific setups. once the abstraction is right it just travels
English
Kalian itu jbjb sehari sampe berapa sih?
Jujur bingung, aku kalo jbjb kaku 🤣
Indonesia
@setyamickala Denger denger mereka menggunakan metode cache ke hard disk gak ke hbm, tapi apapun itu ini bakal jadi impact ke US lab sih
Indonesia
Bro literally Xiaomi sekarang lagi gaspol bikin pricing AI yang gila-gilaan.
MiMo-V2.5-Pro cuma $0.0036 per juta token buat cache hit. Outputnya $0.87.
Sementara Claude masih ngecharge orang kayak lagi jual organ. 😂
Mereka bahkan upgrade credit plan tanpa naikin harga.
Dari yang tadinya cuma ratusan juta credit, sekarang naik berkali lipat.
Lite $6 dapet 4.1B credit. Max $100 dapet 82B.
Sementara Claude?
Masih aja pake pricing "premium experience" sambil kasih rate limit yang bikin orang trauma.
kalian bayar mahal, tapi tiap jam harus nunggu "You've reached your limit".
Orang masih rela bayar $20/bulan buat Claude Pro cuma buat ngerasain "vibe coding yang bagus", padahal sekarang ada provider yang harganya 10x lebih murah dan performanya udah nggak kalah jauh.
walaupun ini china tapi worth it 😂😂😂
Indonesia
cross agent portability is the part i like most. same setup works on claude code, codex, opencode, cursor. as the tool landscape shifts, that flexibility keeps paying off. github.com/affaan-m/ECC
English