TLDR Dev 2026-03-17

As the first engineer on TLDR's new Applied AI team, you'll build AI agents to automate end-to-end workflows and lead our context engineering project to allow non-technical teammates create their own AI workflows. Learn more.

AI agents using MCP often have context bloat, where extensive tool definitions consume tens of thousands of tokens. In these cases, giving your agent a CLI instead is easier, as it allows for progressive discovery of capabilities through `--help` commands. This CLI approach reduces token consumption and improves reliability by running locally, but is worse at org-wide authentication.

Apache Spark's new Real-Time Mode addresses the long-standing trade-off between high-throughput ETL and ultra low-latency streaming workloads. It evolves Spark Structured Streaming's microbatch architecture into a hybrid execution model to achieve millisecond-level latencies. This was done through longer duration epochs with continuous data flow, concurrent processing stages, and non-blocking operators.

FreeBSD has great documentation, performance, and stability compared to Linux. This is why this dev has used FreeBSD for servers and critical tasks with no issues for over twenty years.

AI coding tools are creating a standoff inside software teams. Engineers think they can do product, PMs think they can ship code, and everyone's racing to claim the highest-leverage identity.

Find out how to get leadership comms right in this exclusive webinar. Explore the latest research about authentic leadership communication with experts from Atlassian's Global Communications team and Forrester. Plus, learn from real-world examples of technology leaders using Loom (AI video) to keep customers, investors, and teams aligned - without adding meetings. Join live tomorrow, March 18

Godogen is an AI pipeline with Claude Code skills for building complete Godot 4 game projects from a text description. It orchestrates the entire development process, from designing architecture and generating art (2D/3D) to writing all GDScript code.

Vite+ is a unified, zero-config toolchain to streamline local web development by consolidating runtime, package management, and frontend tools into a single dependency. It integrates popular utilities like Vite, Vitest, Oxlint, and Oxfmt, with commands for development, testing, building, and monorepo task execution.

Oxyde ORM is a type-safe, Pydantic-centric asynchronous ORM for Python, built with a high-performance Rust core for clarity, speed, and reliability. It has a Django-style API for database interactions, supports Pydantic v2 models for full validation, and is designed for modern async Python applications.

Developers are increasingly using AI for learning code. Despite this rising adoption, a "trust gap" persists, as many developers don't have confidence in AI-generated results. This AI tax necessitates a validation step, prompting devs to cross-reference AI-provided information with human-curated resources like technical docs and online communities.

Programming has shifted from building with understood parts to "basic science" on foreign libraries. For example, vibe-coded libraries are foreign as they are built for functionality without real understanding. Still, true mastery and efficiency come from committing to understanding how things really work.

Learn how to set up Sentry's MCP server to pull issues from your projects, trigger Seer's root cause analysis, and apply fixes directly in your IDE. See the recipe

Frustrated by AI-generated code, a developer created a personal skill repository of Node.js best practices and preferences to guide AI assistants in producing higher-quality code.

Deep Agents is a LangGraph-powered, open-source agent harness that provides a ready-to-run AI agent with tools like planning, file access, and shell execution.

This dev attempted to create a single-page feed aggregator for the "small web," only to discover it is unexpectedly vibrant and too active to practically compile all updates.

This dev replaced Google Home with a reliable and private locally hosted voice assistant powered by Home Assistant Assist.