The Bitter Lesson of LLM Extensions

kept by eddie nov 25, 25

The evolution of LLM extensions shows that giving agents general-purpose computing abilities is more effective than specialized tool protocols like MCP.

AddyOsmani.com - Agent Skills
AI coding agents skip senior-engineer practices by default, so Addy Osmani created Agent Skills to encode workflows like specs, tests, and reviews as non-optional agent steps.
also on Agent Skills, Cursor, Claude Code, Anthropic, #software-engineering, #developer-tools
Dynamic context discovery
Cursor introduces dynamic context discovery, pulling only necessary data into the context window via files, reducing agent tokens by up to 46.9%.
also on Agent Skills, Cursor, #software-engineering, #developer-tools
Claude Code and What Comes Next - by Ethan Mollick
Claude Code, using agentic harness and compaction, autonomously built a working website and game, showing a major leap in AI capability for sustained work.
also on Model Context Protocol, Claude Code, Anthropic, #ai-agents, #software-engineering, #ai, #developer-tools
How we contain Claude across products
Anthropic details how it contains AI agent blast radius across three products using sandboxes, VMs, and human-in-the-loop controls, while documenting missed risks like pre-trust code execution and user-mediated prompt injection.
also on Model Context Protocol, Claude Code, Anthropic, #ai-agents, #software-engineering, #developer-tools