Codú — the community for AI builders & indie hackers
How good is AI at grammar checking?
Can you use a large language model for grammar checking without sharing your data with the world?
Codú is growing up
After years as Ireland's largest coding community, here is where it's headed, and why this one is personal for me.
Loops are replacing prompts. Verification is about to be your biggest problem. ↗
Something shifted in the AI coding discourse this month. The argument is no longer about whether agents can write production The post Loops are replacing prompts. Verification is about to be your biggest problem. appeared first on The New Stack.

Markr: Mark the Moment While You Record, Not After ↗
I always wanted to do YouTube but hated the part after recording: rewatching a 20-30 minute video once to find my mistakes, then again to pull shorts out of it. It made me quit more than once. I even tried building an AI auto editor, but it never hit the quality I wanted. So I made Markr instead, a...

How to Make A Clicker Game on the Desmos Graphing Calculator ↗
What if I told you Desmos has everything you need to build a working arcade game? State variables, conditional logic, randomization, and even clickable objects. No JavaScript. No libraries. Just math doing things math was never meant to do.

Parse PDFs for RAG Locally with Docling: Rich Tables, No Cloud Upload ↗
Enterprise Document Intelligence [Vol.1 #5ter] - Table cells, OCR, captions, headings: cloud-grade structure, running on your own machine. No key, no per-page bill, nothing leaves the building The post Parse PDFs for RAG Locally with Docling: Rich Tables, No Cloud Upload appeared first on Towards Da...

Fable 5 vs Opus 4.8: The real stakes, not the spec sheet ↗
This week Anthropic finally dropped the first model in its Mythos-class tier, Fable 5. The pitch was clear: this is The post Fable 5 vs Opus 4.8: The real stakes, not the spec sheet appeared first on The New Stack.

Stop Slicing Your Text Like Salami: A Better Approach to Semantic Chunking ↗
You are ruining your vector search by blindly slicing text into arbitrary character limits. This article explains why standard chunking fails and provides a runnable, dependency-free Python script for context-aware sentence grouping that you can test right now.

TextGrad vs. DSPy & ProTeGi: Evolution of Textual Autograd ↗
Compare TextGrad to prompt optimization frameworks like DSPy and ProTeGi. Discover how textual backpropagation unlocks instance optimization for LLM agents.

Agentic Radiotherapy Planning: Automating Outer-Loop Tuning via TextGrad ↗
Discover how TextGrad automates radiotherapy treatment planning. Learn how language model feedback optimizes hyperparameter weights for precise tumor targeting and tissue sparing.

Solving the 3Blue1Brown String Probability Problem (Without AI) ↗
Let's practice data science thinking through a probability problem The post Solving the 3Blue1Brown String Probability Problem (Without AI) appeared first on Towards Data Science.

Bybit Lists Tokenized SpaceX on Spot as the Trillion-Dollar IPO Lands on Crypto Rails ↗
SpaceX listed on Nasdaq at $135 per share on June 12, 2026, the largest IPO in history at a $1.77 trillion valuation. The stock closed +19% at $160.95 after touching $175.80 intraday. Bybit listed SPCXX, tokenized SpaceX exposure under the xStocks framework, on its Spot market at 16:20 UTC the same...

Claude Fable cost $9 in one coding test. GPT-5.5 cost $1.50. Model triage is the new AI skill. ↗
I’m Matt Burns, Chief Content Officer at Insight Media Group. Each week, I round up the most important AI developments, The post Claude Fable cost $9 in one coding test. GPT-5.5 cost $1.50. Model triage is the new AI skill. appeared first on The New Stack.

Mastra vs LangChain: Building an AI Agent Pipeline and Analyzing the Data ↗
A week ago, I saw this tweet: I had just shipped SupportMesh, a multi-tenant AI support platform built on Mastra, so I had opinions from production. I liked the .dowhile() loop, the typed step schem

How Large-Scale Platforms Handle Millions of Daily Transactions ↗
Every day, millions of people order food, stream videos, send messages, book rides, make payments, and shop online. Most of these actions take only a few seconds from the user's perspective. A user cl

The Saga Pattern in Node.js: How to Roll Back Distributed Transactions Across Microservices ↗
Building reliable workflows across multiple microservices is challenging. In a monolith, a database transaction can ensure that multiple operations either succeed or fail together. But once data is sp

How to Build a Browser-Based PDF Crop Tool Using JavaScript ↗
PDF files often contain unwanted margins, blank spaces, scanner borders, page headers, page footers, or unnecessary content around the main document area. Cropping allows users to remove these unwante

Geopolitical Risk Isn't One Thing. I Built a Python Framework to Prove It ↗
On April 3, 2025, the US announced sweeping tariffs on Chinese imports. SPY dropped 4.8% that day. The next day, it dropped another 6%. Financial news ran the usual headline: markets rattled by geopol

The Hidden Cost of Manual Compliance Testing - And the 60% Time Reduction We Found ↗
The hours-saved business case will be enough to get approved. Use it. But you’ll see other benefits later, and you should track them so the next person doesn’t have to relitigate the case from scratch.

Xbox as a Platform: Is It Time to Face the Truth? ↗
Xbox's future is shrouded in uncertainty. Could we be looking at the end of its console era?

Federal government orders Anthropic to pull Fable 5 and Mythos 5, three days after launch ↗
Anthropic abruptly disabled Fable 5 and Mythos 5 for all customers Friday evening, after the US government issued an export The post Federal government orders Anthropic to pull Fable 5 and Mythos 5, three days after launch appeared first on The New Stack.




