A developer detailed the process of running Qwen3.8-27B on dual RTX 3090s without NVLink under WSL2, achieving 170-210 tok/s on code/JSON. The setup required specific CUDA 13.0 toolchain, SGLang 0.5.1…
A fork of antirez/ds4 optimizes DeepSeek V4 Flash for NVIDIA GeForce RTX 5080 with 16 GB VRAM, achieving a 7.6% decode speedup (3.687 tok/s vs 3.427 tok/s baseline) and a 49.2% prefill improvement (49…
Qubitz, a local-first AI agent for GGUF models on llama.cpp, aims to make 7B–35B MCP/tool-capable LLMs more predictable and useful through a specialized harness and Agent Behavioral Contracts. It oper…
VLLM Launcher, a Windows desktop app for managing local LLM inference through WSL2, has been released as v0.1.0-alpha. The app supports multiple engines including vLLM, SGLang, and llama.cpp, and prov…
Two autonomous AI agents governed a portfolio of 21 software projects for 21 weeks (152 days) using only a shared git repository as the coordination layer, with no message broker, shared database, or …
A developer has built a workflow that runs every untrusted repository through a local LLM triage pass before opening it in an editor, after encountering a fake-recruiter take-home project that contain…
A developer compared Qwen2.5-Coder and DeepSeek-Coder for Solidity smart contract review, running both models locally via Ollama on WSL2 at 7B and below. The developer found that Qwen2.5-Coder followe…
A developer describes keeping an SSH deploy key exclusively inside an encrypted password manager vault, never as a file on disk, after a near-miss where a private key was pasted into an AI chat window…
A second-year AI and Data Science student built an observable AI assistant using SigNoz, OpenTelemetry, Groq, and Streamlit for the Agents of SigNoz hackathon. Despite having no prior experience with …
A developer with 18 years of software experience, including 8 in crypto, is entering their first Sherlock audit contest. The engineer, who built the open-source AI smart contract auditor spectr-ai, de…
A developer who published 67 posts on Dev.to in three months found that practical how-to guides, such as a local Ollama setup, consistently outperformed deep security content by a factor of five in vi…
A developer solved the problem of Claude Code not working with Chrome in WSL2 by using Chrome DevTools Protocol (CDP) with a single Chrome instance on port 9222 and two drivers: one for cheap accessib…
A developer set up a local AI coding agent using Ollama on Windows and Aider inside Ubuntu on WSL2, running the Qwen2.5-Coder model entirely on the local machine without a dedicated GPU. The architect…
Poindexter developers fixed a memory leak in cadvisor that caused an OOM cascade in the WSL2 VM, and patched a content generation bug where LLMs were pulling audit logs and session transcripts as cont…
AutoGPT has evolved from a standalone autonomous script into a low-code platform for building, deploying, and managing continuous AI agents. The platform includes a block-based agent builder, a server…
A developer warns that connecting an MCP server to an AI coding agent introduces two distinct security threats: the agent executing destructive commands and the agent being manipulated by untrusted da…
A developer created a systemd user service to bridge Bitwarden Desktop's SSH agent from Windows to WSL2 using socat and npiperelay, solving hanging issues with ssh-add. The solution requires the stand…
Linux 7.1 was released on June 14, 2026, featuring a new NTFS driver with up to 110% faster multi-threaded writes, but the release cycle was nearly derailed by an influx of AI-generated patches that o…
A developer's setup guide for the Dell XPS 16 9640 and ThinkPad P14s Gen 6 highlights key optimizations for WSL2, Docker, and thermal management. The guide advises limiting WSL2 resources to prevent R…
A developer has published a guide for running Google's Gemma-4 12B instruction-tuned model on Windows Subsystem for Linux 2 (WSL2) using the llama.cpp framework. The process involves installing build …