Measuring AI coding proficiency by token usage is ineffective, as both skilled and struggling developers consume tokens, but skilled users achieve more with fewer tokens, sharper prompts, and better p…
An experiment using Claude Code to create a multi-agent AI workflow for software development, orchestrated through a Java/Spring Boot backend. The author found that the main challenges were not code g…
Arthur Jean built Paneflow, a native terminal workspace for AI coding agents, using Rust and the GPUI framework. The tool is designed to address the limitations of traditional terminal multiplexers by…
The team replaced their chaotic, individualistic use of AI coding tools with a structured methodology called Spec Anchor Development. This approach uses persistent "spec anchors" tied to specific feat…
11 AI plugins for Godot in 2026, categorizing them by how they integrate with the editor (in-editor agents, MCP bridges, AI-native editors, and external clients). After running each through identical …
A documented user investigation from late April 2026 found that on one Windows machine, only 69 of 715 Claude Code sessions (roughly 10%) were visible in the desktop application's sidebar, while the r…
A system called "Minions," inspired by a Stripe blog post, which uses a markdown-based orchestrator to run multiple AI coding agents in parallel on independent tasks. Each agent operates in an isolate…
A new Claude skill called "spec-driven-development" that creates standardized specification files (requirements.md, design.md, tasks.md) and matching configuration files for multiple AI coding tools (…
DEPOSE, a cryptographic forensic framework designed to create verifiable audit trails for AI coding agent sessions (like Claude Code or Codex CLI). It addresses the problem that standard session logs …
Methodology for testing distributed and stateful systems using AI coding agents that execute two specialized skills: one to design a claim-driven test plan and another to execute it. The approach prod…
Multi-agent AI orchestration fails without proper configuration, comparing it to Olympic runners in a relay race who don't know the order or handoff protocol. It details how different tools like Claud…
In an experiment detailed on the AfterPack blog, an AI agent (Claude Code) successfully deobfuscated two JavaScript files—one from an open-source obfuscator and one from a commercial enterprise produc…
An "agentic doc harness" for Obsidian that uses Claude Code skills to navigate notes through wikilinks rather than vector RAG retrieval, achieving higher scores on faithfulness, grounding, insight nov…
The author ran three Claude Code sub-agents (an explore-reviewer, security-reviewer, and plan-architect) on the same 500-line pull request and found that the agents disagreed on 41% of their 78 total …
Figma has launched a new AI agent within its collaborative canvas that allows users to generate and edit designs, automate tasks, and run multiple agents simultaneously using natural language prompts.…
AI coding assistants have not democratized software development but instead widened the skill gap between junior and senior engineers. While both can produce code quickly, a senior developer’s deep ex…
"konid," a language-learning MCP server built by the author after six months of translating Japanese work emails. Unlike standard translation tools, konid returns three options per query ordered from …
To build an AI agent using Claude Code and Zernio that automatically analyzes video or image content, generates platform-optimized captions and hashtags, and publishes to nine social media platforms s…
`dolphindb-agent-skills`, a tool that can be installed via pip in under 30 seconds to provide AI coding assistants like Claude Code and Cursor with an offline knowledge base for the DolphinDB database…
Rise of "cloud coding agents," a new class of AI development tools that execute tasks on remote servers rather than a user's local machine. Unlike traditional IDE assistants that require real-time int…