llm-echo 0.3
Release: llm-echo 0.3 (https://github.com/simonw/llm-echo/releases/tag/0.3) • Mechanisms for testing tool calls (https://github.com/simonw/llm-echo/blob/0.3/README.md#tool-calling). #3 •
Simon Willison's Weblog
datasette-files 0.1a3
Release: datasette-files 0.1a3 (https://github.com/datasette/datasette-files/releases/tag/0.1a3) I'm working on integrating datasette-files into other plugins, such as datasette-extract
Quoting Georgi Gerganov
Note that the main issues that people currently unknowingly face with local models mostly revolve around the harness and some intricacies around model chat templates and prompt construction.
datasette-llm 0.1a3
Release: datasette-llm 0.1a3 (https://github.com/datasette/datasette-llm/releases/tag/0.1a3) Adds the ability to configure which LLMs are available for which purpose
llm-mrchatterbox 0.1
Release: llm-mrchatterbox 0.1 (https://github.com/simonw/llm-mrchatterbox/releases/tag/0.1) See Mr. Chatterbox is a (weak) Victorian-era ethically trained model you can run on your own computer
Mr. Chatterbox is a (weak) Victorian-era ethically trained model you can run on your own computer
Trip Venturella released Mr. Chatterbox (https://huggingface.co/tventurella/mr_chatterbox_model), a language model trained entirely on out-of-copyright text from the British Library. Here's how he
Pretext — Under the Hood
Tool: Pretext — Under the Hood (https://tools.simonwillison.net/pretext-explainer) See my notes on Pretext here (https://simonwillison.net/2026/Mar/29/pretext/).
Pretext
Pretext (https://github.com/chenglou/pretext) Exciting new browser library from Cheng Lou, previously a React core developer and the original creator of the react-motion
Python Vulnerability Lookup
Tool: Python Vulnerability Lookup (https://tools.simonwillison.net/python-vulnerability-lookup) I learned that the OSV.dev (https://osv.dev/) open source vulnerability database has an open CORS
Quoting Matt Webb
The thing about agentic coding is that agents grind problems into dust. Give an agent a problem and a while loop and - long term - it’ll solve that problem even if it means burning a trillion
datasette-showboat 0.1a2
Release: datasette-showboat 0.1a2 (https://github.com/simonw/datasette-showboat/releases/tag/0.1a2) I added an option to export a Markdown file from my app that lets Showboat incrementally
Quoting Richard Fontana
FWIW, IANDBL, TINLA, etc., I don’t currently see any basis for concluding that chardet 7.0.0 is required to be released under the LGPL. AFAIK no one including Mark Pilgrim has identified
Vibe coding SwiftUI apps is a lot of fun
I have a new laptop - a 128GB M5 MacBook Pro, which early impressions show to be very capable for running good local LLMs. I got frustrated with Activity Monitor and decided to vibe code up some
We Rewrote JSONata with AI in a Day, Saved $500K/Year
We Rewrote JSONata with AI in a Day, Saved $500K/Year (https://www.reco.ai/blog/we-rewrote-jsonata-with-ai) Bit of a hyperbolic framing but this looks like another case study of vibe-porting, this
My minute-by-minute response to the LiteLLM malware attack
My minute-by-minute response to the LiteLLM malware attack (https://futuresearch.ai/blog/litellm-attack-transcript/) Callum McMahon reported the LiteLLM malware attack
Quantization from the ground up
Quantization from the ground up (https://ngrok.com/blog/quantization) Sam Rose continues his streak (https://simonwillison.net/tags/sam-rose/) of publishing spectacularly informative interactive
datasette-files-s3 0.1a1
Release: datasette-files-s3 0.1a1 (https://github.com/datasette/datasette-files-s3/releases/tag/0.1a1) A backend for datasette-files (https://github.com/datasette/datasette-files) that adds the
Thoughts on slowing the fuck down
Thoughts on slowing the fuck down (https://news.ycombinator.com/item?id=47517539) Mario Zechner created the Pi agent framework (https://github.com/badlogic/pi-mono) used by OpenClaw, giving
datasette-llm 0.1a1
Release: datasette-llm 0.1a1 (https://github.com/datasette/datasette-llm/releases/tag/0.1a1) New release of the base plugin that makes models from LLM (https://llm.datasette.io/) available for
LiteLLM Hack: Were You One of the 47,000?
LiteLLM Hack: Were You One of the 47,000? (https://futuresearch.ai/blog/litellm-hack-were-you-one-of-the-47000/) Daniel Hnyk used the BigQuery PyPI dataset