Trace Sampling for LLM Apps: Keep the Spans That Matter, Drop the Rest
Storing every LLM trace at full fidelity gets expensive fast. Here is a sampling policy that keeps the errors, the slow calls, and the eval set.
Tutorials
Why Infrawise Uses Deterministic Analysis Instead of an LLM
Ask your AI coding assistant which Global Secondary Indexes exist on your Orders table. It will read...
The billing bet that killed my coding-agent harness
Filomena is a local harness I built to run a few Claude Code agents from one terminal. Six months of...
JSON or XML Tags for LLM Output: The Format That Holds Under Pressure
JSON for the payload, XML tags for the wrapper. Which format survives nesting, partial streams, and a model that wandered off-schema.
Metadata Filtering Before Vector Search: The Recall Win Nobody Measures
Pre-filter by metadata to shrink the search space before the vector index runs. The recall lift, the cardinality trap, and the code.
The Paradox of Power: Why Anthropic Released and Then Restricted Claude Fable 5
On June 9, 2026, Anthropic released Claude Fable 5, which was described as the most capable AI model...
Bun rewrote itself from Zig to Rust in 9 days with an LLM. That's terrifying.
It took nine days to rewrite all of Bun from Zig to Rust using an LLM and get the new code to the...
I Spent $200 in One Day on AI Tokens to Build 86 Browser Tools — A Solo Founder's Complete Build Log
86 tools. Zero backend. One person. Two hospital visits. $200 in tokens in 24 hours. This is what...
I've Used Cursor Every Day for Months — These 5 Habits Did More Than Any Setting
I see a lot of posts hunting for the secret Cursor setting that unlocks 10x productivity. After using...
I almost gave up on my AI assistant — here’s how I fixed context handling
I’ve been building a personal AI assistant for the past few months. You know the kind: you chat with...
AI should do the implementation. You should own the decisions.
Son of Anton is a delivery orchestrator for solo devs and small teams. Three human gates; the orchestrator owns everything in between. Runs on any agent that reads AGENTS.md.
Build Your Own Shakespearean LLM
You know about LLMs (Large Language Models), but how are they created? Let's build our own to find...
I Reach for Cursor 90% of the Time — Here's the 10% Where Claude Code Wins
Most of the "Cursor vs Claude Code" takes I read are framed wrong. It's not a cage match. They're not...
AI-Native Data Engineering: From ETL Pipelines to Agentic Data Serving
TL;DR Traditional decoupled ETL pipelines (like the "Modern Data Stack") are too brittle...
System Architect vs. AI Solution Architect: An Anatomy of Roles
I've observed fundamental differences between struggling with the overall stability and performance...
The day a refactor passed on my laptop and failed on yours
Most of the code being written right now is not being written. It is being generated, glanced at, and...
My AI System Logged 35,669 LLM Calls. It Still Couldn’t Tell Me What They Cost.
CORE had telemetry. That was the comforting part. Every LLM exchange was being logged. Prompt...
Shipping an Android podcast player with on-device Whisper — 6 lessons from a Czech indie launch
After a year of nights and weekends, I shipped Lucidcast — an Android podcast player I built because...
First impressions of Fable 5 and the racing game I built with it
My first impressions of Fable 5 before it was suspended. Best way to use Fable 5 is to start with...
How to Build Autonomous AI Agent Skills for Claude Code
A practical guide to building effective Claude Code skills, from guardrails and workflows to reflection loops and token economics. Includes real experiment results.
NeuralBridge: Self-Healing SDK for LLM-Powered AI Agents - Getting Started in 5 Minutes
NeuralBridge is an embedded self-healing SDK for LLM-powered AI agents. pip install in seconds, zero infrastructure.
Playwright MCP: cómo dar navegador a un agente de IA sin convertir tus tests en una caja negra
Playwright MCP permite que un agente controle un navegador real, pero no reemplaza una suite de...
KeepAI: a local, open-source API hub that lets AI agents use your apps safely
AI agents are getting good at doing things — triaging your inbox, updating a Notion doc, opening a...
AI Agent Skills Are Everywhere. But Which Ones Are Safe to Install?
AI agents are becoming less like single chatbots and more like extensible runtimes. Instead of using...