AI Dev Stack

Sign in Subscribe

feature

Anthropic's Hiring Test Got Beat by Claude — Twice

Anthropic's Hiring Test Got Beat by Claude — Twice

Anthropic's performance engineering take-home has been redesigned three times because each new Claude model solved it. The latest version uses deliberately weird puzzles to stay ahead of AI — because realism is now a luxury they can't afford.

16 Parallel Claudes Built a C Compiler That Boots Linux

16 Parallel Claudes Built a C Compiler That Boots Linux

Anthropic researcher ran 16 parallel Claude instances for 2 weeks to build a 100,000-line C compiler that boots Linux. No human intervention. $20k in API costs. This is what autonomous agents can barely do today — and what they'll do routinely tomorrow.

Contextual Retrieval: Anthropic's Fix for RAG's Biggest Problem

Contextual Retrieval: Anthropic's Fix for RAG's Biggest Problem

Anthropic's Contextual Retrieval uses Claude to add context to document chunks before embedding, cutting RAG retrieval failures by 49%. Combined with reranking, it hits 67%. Here's how it works and why it matters.

Building Effective AI Agents: Anthropic's Production Patterns

Building Effective AI Agents: Anthropic's Production Patterns

Anthropic's internal playbook for building production AI agents: five workflow patterns, when to use autonomous agents vs predefined workflows, and why tool design matters as much as prompts.

Claude 3.5 Sonnet Hits 49% on SWE-Bench: How Anthropic Built It

Claude 3.5 Sonnet Hits 49% on SWE-Bench: How Anthropic Built It

Claude 3.5 Sonnet hit 49% on SWE-bench Verified with a minimal agent scaffold: two tools, one prompt, maximum model control. Here's the exact architecture Anthropic used and why tool design matters as much as model capability.

Claude's "Think" Tool: When to Stop and Reason Mid-Task

Claude's "Think" Tool: When to Stop and Reason Mid-Task

Anthropic's "think" tool gives Claude structured reasoning checkpoints during execution. It delivered 54% improvement on policy-heavy tasks and 1.6% on SWE-Bench. Here's when to use it—and when to skip it.

Claude Code: Anthropic's Full-Stack AI Coding Assistant

Claude Code: Anthropic's Full-Stack AI Coding Assistant

Anthropic's Claude Code runs in terminal, IDE, browser, and mobile with session portability across all surfaces. Same context, different environments.