Flux
Toutes les catégories

Programmation

1004 articles

Claude system prompts as a git timeline

Claude system prompts as a git timeline

Research: Claude system prompts as a git timeline Anthropic publish the system prompts for Claude chat and make that page available as Markdown. I had Claude Code turn that page into separate files for each model and model family with fake git commit dates to enable browsing the changes via the GitHub commit view. I used this to write my own detailed notes on the changes between Opus 4.6 and 4.7. Tags: system-prompts, anthropic, claude, generative-ai, ai, llms

Simon Willison's Weblog
Adding a new content type to my blog-to-newsletter tool

Adding a new content type to my blog-to-newsletter tool

Agentic Engineering Patterns > Here's an example of a deceptively short prompt that got a lot of work done in a single shot. First, some background. I send out a free Substack newsletter around once a week containing content copied-and-pasted from my blog. I'm effectively using Substack as a lightweight way to allow people to subscribe to my blog via email. I generate the newsletter with my blog-to-newsletter tool - an HTML and JavaScript app that fetches my latest content from this…

Simon Willison's Weblog
Join us at PyCon US 2026 in Long Beach - we have new AI and security tracks this year

Join us at PyCon US 2026 in Long Beach - we have new AI and security tracks this year

This year's PyCon US is coming up next month from May 13th to May 19th, with the core conference talks from Friday 15th to Sunday 17th and tutorial and sprint days either side. It's in Long Beach, California this year, the first time PyCon US has come to the West Coast since Portland, Oregon in 2017 and the first time in California since Santa Clara in 2013. If you're based in California this is a great opportunity to catch up with the Python community, meet a whole lot of interesting people…

Simon Willison's Weblog
Socket Named Top Sales Organization by RepVue

Socket Named Top Sales Organization by RepVue

Socket has been named a 2026 Reppy Award recipient by RepVue in two categories: Small Companies and Venture Capital Backed Companies. RepVue is the leading platform for B2B sales reps to rate their own employers, with more than 225,000 users, and Reppys recognize top-rated sales organizations based on employee ratings across categories like Culture & Leadership, Compensation, and Product-Market Fit. Socket earned a RepVue Score of 94.25, placing us in the top 5% of all companies on the…

Socket
PHP Alive And Kicking: Episode 22 Sara Golemon

PHP Alive And Kicking: Episode 22 Sara Golemon

Sara Golemon returns to the show — her first appearance was way back in Episode 2 — for a deep dive into Git Worktrees, a powerful but underused Git feature that lets you work on multiple branches simultaneously without the overhead of stashing, context-switching, or rebuilding from scratch. The episode covers practical real-world use cases, […] The post PHP Alive And Kicking: Episode 22 Sara Golemon appeared first on PHP Architect.

PHP Architect
NIST Officially Stops Enriching Most CVEs as Vulnerability Volume Skyrockets

NIST Officially Stops Enriching Most CVEs as Vulnerability Volume Skyrockets

NIST is moving to a risk-based enrichment model for the National Vulnerability Database, formally abandoning its longstanding goal of analyzing every submitted CVE. Starting immediately, the NVD will only enrich vulnerabilities that appear in CISA's Known Exploited Vulnerabilities (KEV) catalog, software used by the federal government, or software designated as critical under Executive Order 14028. Everything else gets labeled "Not Scheduled." The announcement came during VulnCon, where NVD…

Socket
datasette 1.0a28

datasette 1.0a28

Release: datasette 1.0a28 I was upgrading Datasette Cloud to 1.0a27 and discovered a nasty collection of accidental breakages caused by changes in that alpha. This new alpha addresses those directly: Fixed a compatibility bug introduced in 1.0a27 where execute_write_fn() callbacks with a parameter name other than conn were seeing errors. (#2691) The database.close() method now also shuts down the write connection for that database. New datasette.close() method for closing down all databases and…

Simon Willison's Weblog
The PHP Podcast 2026.04.16 with Sara, Joe, and Sami

The PHP Podcast 2026.04.16 with Sara, Joe, and Sami

The PHP Podcast streams live, typically every Thursday at 3 PM PT. Come join us and subscribe to our YouTube channel. Another fun episode of the PHP Podcast! Here’s what we covered: 🎙️ Sammy Powers Returns! – After 4 years away from the PHP community, Sammy joins us from Germany where he’s working on immigration […] The post The PHP Podcast 2026.04.16 with Sara, Joe, and Sami appeared first on PHP Architect.

PHP Architect
llm-anthropic 0.25

llm-anthropic 0.25

Release: llm-anthropic 0.25 New model: claude-opus-4.7, which supports thinking_effort: xhigh. #66 New thinking_display and thinking_adaptive boolean options. thinking_display summarized output is currently only available in JSON output or JSON logs. Increased default max_tokens to the maximum allowed for each model. No longer uses obsolete structured-outputs-2025-11-13 beta header for older models. Tags: llm, anthropic, claude

Simon Willison's Weblog
Socket Selected for OpenAI's Cybersecurity Grant Program

Socket Selected for OpenAI's Cybersecurity Grant Program

OpenAI has named Socket as one of the initial recipients of its Cybersecurity Grant Program, a new initiative that commits $10 million in API credits to support organizations advancing cybersecurity defense. The grant comes alongside access to more cyber-permissive frontier models through Trusted Access for Cyber, OpenAI's new identity-based framework for defensive acceleration. Both programs select for trusted defenders with a proven track record in identifying and remediating vulnerabilities…

Socket
Qwen3.6-35B-A3B on my laptop drew me a better pelican than Claude Opus 4.7

Qwen3.6-35B-A3B on my laptop drew me a better pelican than Claude Opus 4.7

For anyone who has been (inadvisably) taking my pelican riding a bicycle benchmark seriously as a robust way to test models, here are pelicans from this morning's two big model releases - Qwen3.6-35B-A3B from Alibaba and Claude Opus 4.7 from Anthropic. Here's the Qwen 3.6 pelican, generated using this 20.9GB Qwen3.6-35B-A3B-UD-Q4_K_S.gguf quantized model by Unsloth, running on my MacBook Pro M5 via LM Studio (and the llm-lmstudio plugin) - transcript here: And here's one I got from Anthropic's…

Simon Willison's Weblog
Esc