cd/sources/dev-to· home› sources› Dev.to
cat /sources/dev-to.feed | wc -l → 26841

Dev.to

articles 26841 domain dev.to → page 104/1343 feed RSS
07:32
2026-10-24
dev.to
ai-research

AI Helps Us Build Faster. Are We Learning Enough?

Anthropic researchers ran a January 2026 study in which 52 mostly junior engineers learned the unfamiliar Python library Trio, with the AI-assisted group averaging 50% on a post-task assessment versus…

07:30
2026-10-24
dev.to
agent-protocols

llms.txt v2: what an agent actually reads

A developer tested the August 2026 revision of the llms.txt proposal, running this site's own /llms.txt through the proposal's roughly twenty-line reference parser (llms_txt/miniparse.py) and finding …

07:29
2026-10-24
dev.to
ai-agents

Why I Stopped Letting Claude Code Touch My Main Drive

A developer described switching to a per-drive-letter sandbox on Windows for Claude Code after the agent deleted half a node_modules folder during a refactor. Each agent session is now mapped to its o…

06:06
2026-10-24
dev.to
ai-agents

What my review gate does about weakened or deleted tests

A developer released two-engines-gate, an MIT-licensed toolset in which Claude Code and Codex review each other's commits, and documented a gap where a commit that weakens tests alongside broken code …

06:06
2026-10-24
dev.to
ai-infrastructure

Why Monolithic AI Gateways Are Losing Ground in 2026

A developer's retrospective argues that monolithic AI gateways break down as teams add providers, protocols and voice support, and points to Respan, an open-source AI gateway written in Rust, as a plu…

06:04
2026-10-24
dev.to
ai-agents

Why AI Coding Agents Break on Production Pipelines

An engineer reports that general-purpose AI coding assistants such as GitHub Copilot fail on production automation tasks because they generate code snapshots without modeling runtime behavior, rate li…

06:03
2026-10-24
dev.to
ai-tools

A 1.5B model is a triage signal, not a verdict

A developer built a local code-auditing tool that runs five deterministic checks over trading-strategy repositories and then asks a locally hosted Qwen2.5-Coder-1.5B-Instruct Q4_K_M model, via llama.c…

05:34
2026-10-24
dev.to
ai-agents

Misreported compliance is worse than hallucination

A developer behind the Merlin agent system argues that misreported compliance—agents confidently claiming they followed instructions they did not follow—is more damaging than hallucination because not…

05:31
2026-10-24
dev.to
developer-tools

What if managing files could be as simple as a few clicks?

A developer has released Ceyloro, an open-source cross-platform desktop application that puts file conversion and compression behind an intuitive radial menu. The tool is now open for testing, with th…

05:30
2026-10-24
dev.to
ai-tools

Slow Mode: the laziest way to go outside

A developer built Trailhead, a local-first Streamlit app that generates short outdoor outing cards using live weather and daylight data from Open-Meteo and a locally running Gemma 3 4B model via Ollam…

05:00
2026-10-24
dev.to
ai-agents

Stateless agents cut long‑horizon token usage

A paper on Stateless Language Agents (SLAs) reports that decoupling durable state from the model's active context lets long-horizon agents reach the strongest kernel baseline's final performance while…

← prev page 104 / 1343 next →