13 min read
Claude Managed Agents Now Have Filesystem Memory
Claude Managed Agents got filesystem memory in public beta on April 23, 2026, agents read and write actual…
The Lab · Written in the open · Tagged AI Tools
Lab.
Working notes from the studio. The tools I build, the workflows I keep, the mistakes I paid for.
209 entries · latest 2026-11-04
13 min read
Claude's 200+ Connectors Changed How I Use AI
Claude's connector directory crossed 200+ one-click integrations in April 2026, up from 50 in February
8 min read
Claude Code 2.1.118 Adds Vim Mode, /theme, And Hook MCP
Vim visual mode lands in Claude Code 2.1.118 with v and V selection, plus full operator support over…
8 min read
Claude Code Ultraplan: Plan in the Cloud, Run Anywhere
Claude Code Ultraplan drafts plans in the cloud, lets you review them in a web editor, then runs…
9 min read
Anthropic Confirms Claude Code Was Broken For 7 Weeks
March 4 default reasoning dropped from high to medium for latency, hurt coding accuracy on real tasks
8 min read
5 Figma Plugins AI Still Cannot Replace in 2026
Five Figma plugins that still beat AI in 2026: Stark, Tokens Studio, Autoflow, Design Lint, Figmotion. A craft…
8 min read
The 9 Claude Code Hooks That Audit Every File I Write
How 9 shell-script hooks replace code review in a solo studio: what each one catches, how to write…
10 min read
MCP Servers Are How Claude Actually Talks to Everything
MCP servers expose tools, resources, and prompts to Claude through a tiny JSON-RPC protocol.
10 min read
I Built 3 Production Agents With the Claude Agent SDK in One Weekend
Shipped 3 production agents in one weekend using the Claude Agent SDK, each under 50 lines of glue…
10 min read
The 1M Context Window Actually Changes How I Code
Loaded my entire 12-repo studio into one session and stopped re-explaining context to the model every hour
9 min read
How I Run a 15-Repo Studio From One CLAUDE.md File
I run a 15-repo AI creative studio with one CLAUDE.md file as the brain, not 15 different configs
10 min read
Running LLM Evals in Production: The 2026 Guide
Most LLM apps ship with zero evals because engineers confuse demos with quality