Blog
Articles about AI and work.
How to Use PARA in an AI Second Brain (2026)
Four buckets, filed by your assistant
PARA Method vs Building a Second Brain (2026)
One is a method, one is the whole system
ZCode review: the free GLM-5.2 agent vs Claude Code
Cheaper than Claude Code, and it still forgets your project
The PARA Method: A Complete Guide (2026)
Projects, Areas, Resources, Archives
Claude Fable 5 Skills: Keep the Rigor After July 7
6 free skills, blind-tested on Opus 4.8. Receipts included
Claude Code Skills by Fable 5: They Run in Codex Too
The smart model writes. Your daily model keeps it.
How to build a second brain with an AI agent
The agent does the filing, not you
Updated
How to Switch From ChatGPT to Claude: Memory and Projects
Review memory, keep an archive and rebuild one useful project before adding tools.
Updated
9 best AI executive assistants in 2026
Nine tools tested for executive-grade delegation
ChatGPT Work vs Codex vs Claude Code: What Changed and Which One Should You Use?
OpenAI just turned ChatGPT into a broader work agent. Here is how ChatGPT Work, Codex, Claude Code, Gemini CLI, and Antigravity now fit together.
Context Engineering Research: 2026 Papers and Benchmarks
What ContextBench, SWE-contextbench, AGENTS.md, and long-context studies actually test.
AI for ADHD: 6 Tools and a Practical Work System (2026)
Compare AI assistants for ADHD, then copy a five-step workflow that protects context and reduces decisions.
Updated
Claude Code CLI vs Desktop: Which Surface Fits?
A 50-row feature matrix for Claude Code CLI, Desktop, Dispatch, hooks, MCP, model support, and scheduling.
Best knowledge management software in 2026
Seven team tools judged on AI-native recall
Best AI personal assistant in 2026
Seven assistants tested on real daily admin
Claude Code best practices: clear instructions and verified work
A practical way to use CLAUDE.md, scoped rules, plans, and checks without turning repository context into a dumping ground.
Updated
Loop Engineering Is Easy. Deciding What "Done" Means Is the Whole Job.
Prompt engineering taught us to ask. Context engineering taught us what to feed the model. Loop engineering is the next rung, and the hard part was never the loop. It is deciding what "done" means.
How to use AI for work (a system, not prompts)
Build a system, stop collecting prompts
Best AI tools for consultants in 2026
Seven tools tested across the consulting workflow
Best AI tools for product managers in 2026
Seven tools tested across the PM workflow