Engineering
How Anthropic Contains Claude Agents Across Products
Anthropic describes how it isolates Claude AI agents across claude.ai, Claude Code, and Claude Cowork using a defense-in-depth approach at…
Engineering
Postmortem: What Degraded Claude Code Quality in Early 2026
Anthropic identified and resolved three separate issues that degraded Claude Code quality between March and April 2026: a change to…
Engineering
Build Your Own RAG Pipeline from Scratch with Python and Ollama
This Hugging Face tutorial walks through building a Retrieval-Augmented Generation system from scratch using Python and Ollama, demonstrating how RAG…
Engineering
KV Caching Explained: Optimising Transformer Inference Efficiency
Key-Value caching is a technique that accelerates transformer-based text generation by storing intermediate attention states — specifically keys and values…
Engineering
Mastering Tensor Dimensions in Transformers
This Hugging Face article traces how tensor shapes flow through each layer of a decoder-only transformer, from tokenisation and embedding…
Engineering
Building Moon Bot: A Slack-Native Coding Agent with Persistent Memory
Hugging Face built Moon Bot, a Slack-integrated coding agent that consolidates database queries, code navigation, and GitHub operations into a…
Engineering
Eval Awareness: When Claude Opus 4.6 Figured Out It Was Being Tested
Anthropic researchers discovered that Claude Opus 4.6 exhibited ‘eval awareness’ when tested on BrowseComp, independently hypothesizing it was being evaluated,…
Engineering
Building a C Compiler with a Team of 16 Parallel Claudes
Anthropic researcher Nicholas Carlini orchestrated 16 parallel Claude instances to autonomously build a 100,000-line C compiler capable of compiling Linux…
Engineering
Quantifying Infrastructure Noise in Agentic Coding Evals
Anthropic researchers discovered that infrastructure configuration significantly impacts agentic coding benchmarks, with resource allocation alone causing score variations of up…
Engineering
How Anthropic Built Claude Code Auto Mode: A Safer Way to Skip Permissions
Anthropic introduced Claude Code auto mode, a safety system that automates permission decisions to reduce approval fatigue while maintaining security.…
Engineering
Scaling Managed Agents: Decoupling the Brain from the Hands
Anthropic introduced Managed Agents, a hosted service that separates agent components — the brain (Claude and its harness), hands (sandboxes…
Engineering
Harness engineering for coding agent users
Thoughtworks Distinguished Engineer Birgitta Böckeler explores how developers can build trust in AI coding agents by designing an effective “outer…