TL;DR
Have coding agents finally crossed the quality threshold? A five-minute tour of six months in LLMs, Pydantic's hidden costs, and turbovec out-searching FAISS in 4GB of RAM.
🦄
Articles & Tutorials
× 10 01 Batching API calls
02 Shrink Your Python Container in One Command with SlimToolkit
03 Build a Live Object Detection App for the Reachy Mini With TensorFlow and PyCharm
04 When Python Manual Wiring Turns Into Copy-Paste Architecture
05 In-Kernel Broadcast Optimization: Co-Designing Kernels for RecSys Inference
06 The Last Six Months in LLMs in Five Minutes
07 Transformer From Scratch
08 A History of IDEs at Google
09 The Boundary of Pydantic
10 What We Maintain When Maintaining Model API
🐿️
Projects & Resources
× 12 01 RAMPART: Risk Assessment & Measurement Platform for Agentic Red Teaming
02 ZOZO's Contact Solver: A contact solver for physics-based simulations involving shells, solids and rods
03 BleachBit: Cleans files to free disk space and to maintain privacy
04 turbovec: A 10M document vector index that fits in 4GB RAM and searches faster than FAISS
05 Anthropic Cybersecurity Skills: The largest open-source cybersecurity skills library for AI agents
06 LongLive: Real-time Interactive Long Video Generation
07 OBLITERATUS: Advanced open-source toolkit for understanding and removing refusal behaviors from LLMs
08 Webwright: A Terminal Is All You Need For Web Agents
09 whichllm: Find the best local LLM that actually runs on your hardware
10 OmniParse: Ingest, parse, and optimize any data format for enhanced compatibility with GenAI frameworks
11 Kami: Good content deserves good paper
12 Open Design: The open-source alternative to Claude Design