Simon Willison

Person. Independent developer and one of the most-read practical voices on LLMs, blogging prolifically at simonwillison.net. Creator of Datasette (open-source data exploration/publishing) and the LLM CLI tool, co-creator of the Django web framework, and author of a stream of small sharp tools (sqlite-utils, shot-scraper).

His AI writing is hands-on and empirical — testing every model release the day it drops, documenting what actually works. Coined the term "prompt injection" and has been its most persistent chronicler. A strong advocate of open-source tooling around LLMs and of writing things down; his weblog doubles as a running lab notebook for the field.

In this corpus he appears both as a source (posts from his blog) and as a frequently-cited authority in others' writing (e.g. his "Skills over MCP" take in Claude Code practitioner guides).

entities · 5 notes linked

Related: Anthropic · Claude Code · AI Agents · Model Context Protocol · Agentic Coding · Large Language Models · Clip · Drew Breunig

Notes

• Claude Skills are awesome, maybe a bigger deal than MCP — Claude Skills as lightweight Markdown-based agent capability extension beating MCP
• Finding Bathroom Faucets with Embeddings — Using CLIP embeddings to build image-similarity product navigation tool
• How I Use Every Claude Code Feature — Practitioner's guide to Claude Code features and workflows
• NotebookLM's automatically generated podcasts are surprisingly effective — Google NotebookLM generates convincing AI podcast episodes from user documents
• Simon Willison (@simonw) on X — Simon Willison asks for real-world LLM fine-tuning commercial success stories