~/bookmarks

Tiptap is a suite of content editing & real time collaboration tools. Build editor experiences like Notion in weeks, not years.

Post-training with Tinker. Contribute to thinking-machines-lab/tinker-cookbook development by creating an account on GitHub.

Install Openship on macOS, Windows, Linux, or grab the CLI. Native desktop app and command-line - same backend, same deploys, your choice of surface.

Open Source Captable Worksheet . Contribute to 1984vc/cap-table development by creating an account on GitHub.

503 lessons. 20 phases. Write the backprop, the tokenizer, the attention mechanism, and the agent loop by hand before any framework gets imported. Python, TypeScript, Rust, Julia.

An interactive guide to inference engineering: how generative AI models are served in production, from attention kernels to multi-cloud capacity.

We’re on a journey to advance and democratize artificial intelligence through open source and open science.

Public collection of high-quality templates created by General Legal attorneys with CC0 license - General-Legal/legal-templates

Agents get computers now, and definitions vary. A Fly.io Sprite gives each agent its own hardware-isolated Linux machine that keeps its filesystem, stays reachable, and sleeps when there is nothing to do.

The structured context layer for AI agents. Build self-learning specialised agents that are context aware. - DrDroidLab/open-index

The strictest and most opinionated python linter ever! - wemake-services/wemake-python-styleguide

VS Code in the terminal. Contribute to zenbu-labs/terminal-code development by creating an account on GitHub.

Unsloth Dynamic 3.0 GGUFs are post-training quantizations (no QAT/QAD) using a refined imatrix calibration dataset and per-layer quantization selection, compatible with llama.cpp and Unsloth Desktop. Benchmarks for Qwen3.8-27B report over 10% better top-1% accuracy at equal file size versus other providers, evaluated via Divergence-300 @32 and KL Divergence on held-out prompts. File sizes range from UD-IQ1_S at 6.2GB to UD-Q2_K_XL at 9.83GB, with quants under 8.37GB omitting the MTP module to save roughly 500MB.

Documentation for fx, the CLI-based AI coding agent from Vercel Labs.

We’re on a journey to advance and democratize artificial intelligence through open source and open science.

We’re on a journey to advance and democratize artificial intelligence through open source and open science.

Nicholas Charriere, a YC alum with two kids, examines the conflict between high career ambition and active fatherhood, noting that many prominent founders neglected their children. He redefines ambition to include being an engaged dad and outlines a strategy of rigid time boundaries, health focus, and eliminating low-value activities. The post argues that ambitious parents must maximize hours with their children, especially during early years that overlap with career peak.

Qwen 3.8 27B ROCmFP4 on AMD Strix Halo (Ryzen AI Max+ 395). Up to 36 tok/s via MTP Speculation, TurboQuant & Mesa RADV Wave64. - julianmb/q38rocm

Piro engineers what AI tells the world about your brand. A Hollywood storytelling studio making content that AI engines trust, cite, and recommend.

Deploy Qwen3.8-27B with SGLang — dense hybrid GDN vision-language model with BF16/FP8/NVFP4 W4A4 checkpoints and in-checkpoint MTP, single-GPU on H200, RTX PRO 6000, RTX 5090 and DGX Spark.

A collection of notebooks/recipes showcasing some fun and effective ways of using Claude. - anthropics/claude-cookbooks

Taste Skill gives your AI agent good taste. Open-source skill files that stop AI from generating generic, boring frontends.

Anti-laziness skill for AI agents. Core: the Depth Tree method, which splits a task N layers deep and gives every leaf the full time budget of the whole task, so effort multiplies with depth. Groun…

Modular prompts I use every day. Contribute to staltz/microskills development by creating an account on GitHub.