Blog

Thoughts on AI-assisted development, developer workflows, and building tools that stay out of your way.

1. PREDICTED --- predicted: tasks: 3 risk: "seam drift" confidence: 0.8 preconditions: - claim: grep-checkable status: pending frozen at shape time graded 2. GRADED ## verify delta: 0 misses: 0/58 n: 10 confidence ordered: no — size did verdict: 3. FINDING H1: preconditions predict failure not supported at n=10 what bit instead: seam mismatch, unstated env state did the forecast hold? 10 graded specs · 4 PM wave 1 + 6 MM dogfood
Project Management for Developers8 min read

What Grading 10 AI Specs Taught Us About Decomposition

We graded our own AI spec forecasts against outcomes on ten specs: task counts matched every time, precondition checks held 58 of 58, one hypothesis failed.

Read article
Project Management for Developers8 min read

Why Every AI Coding Task Should Start With a Brief

A brief records what you asked for, verbatim, before any agent interprets it. Why Monday Morning makes the brief the first artifact of every coding task.

Justin Young·September 15, 2026
AI Development6 min read

Claude Code CLI vs the API: Why the CLI Wins for Dev

Claude Code CLI vs the Anthropic API for development work — what each is for, why the CLI fits real coding, and how Monday Morning builds on top of it.

Angela Edmundson·June 13, 2026
Project Management6 min read

How Claude Plan Mode Keeps Projects on Track

Claude Code plan mode separates thinking from doing so agents stop running off in the wrong direction. How it works and how Monday Morning builds on it.

Angela Edmundson·June 13, 2026
Industry Trends6 min read

From IDE to ADE: The Agent Development Environment

Development is shifting from single-window IDEs to multi-terminal agent development environments (ADEs). What changes, and where Monday Morning fits.

Angela Edmundson·June 13, 2026
Project Management6 min read

Living Roadmaps: Plans That Evolve With Your Code

In AI development, plans change daily. A living roadmap reconciles with reality as you ship while keeping the birds-eye view intact. How to build one.

Angela Edmundson·June 13, 2026
AI Development6 min read

The Coordination Problem in Terminal AI Agents

Running many AI coding agents in terminals is easy. Coordinating them is not. Why agentic dev needs an orchestration layer — and how the Conductor solves it.

Angela Edmundson·June 13, 2026
Industry Trends7 min read

Vibe Coding vs Building Real Projects With AI

Vibe coding ships demos. Building real projects ships products. The difference for solo founders, agencies, and freelancers — and the tooling that bridges it.

Angela Edmundson·June 13, 2026
MCP & Tool Ecosystem8 min read

What Are MCP Servers and Why They Matter for AI-Assisted Development

MCP servers connect AI coding assistants to your tools and project data. Learn how the Model Context Protocol works and why it changes everything.

Angela Edmundson·May 17, 2026
AI Development4 min read

What Is Context Recovery and Why Every AI-Assisted Developer Needs It

AI coding assistants forget everything between sessions. Context recovery solves this by giving your AI the project state it needs to be productive from the first prompt.

Angela Edmundson·May 17, 2026
Project Management9 min read

Spec-Driven Development: How to Give Your AI Coding Assistant a Blueprint

Spec driven development produces dramatically better AI-generated code. Learn how writing specs before prompting gives AI assistants the structure they need.

Angela Edmundson·May 15, 2026
Project Management8 min read

Why Local-First Project Management Beats Cloud Tools for Developer Teams

Local first project management keeps tasks, specs, and issues in your repo as markdown. Faster, private, offline-capable, and version controlled.

Angela Edmundson·May 13, 2026
AI Development8 min read

How to Build Custom Slash Commands for Claude Code

Learn to build custom Claude Code slash commands using markdown files. Practical guide with examples, variables, and tips for automating dev workflows.

Angela Edmundson·May 11, 2026