$ ls posts/
Posts
- How I Run 8 Claude Code Sessions at Once
The exact workflow for running 6–8 parallel Claude Code instances across different projects. Terminal setup, task scoping, and how I review it all without losing my mind.
workflowclaude-codeparallel - Claude Code Best Practices for Real Projects
10 concrete practices for using Claude Code on production codebases. Not toy projects, not tutorials — the stuff that actually matters when your code ships to users.
claude-codebest-practicesworkflow - The Spec That Makes AI Agents Work Independently
What a spec needs to contain for Claude Code to build without you watching. The difference between a prompt and a specification, with real examples.
specsmethodologyclaude-code - Spec Coding: The Complete Guide
The full methodology for writing specifications that AI agents can execute independently. Where it came from, how it works, and why it replaced vibe coding in my workflow.
spec-codingmethodologyguide - Scoping Tasks for Background Claude Code Agents
How to break down work into tasks that Claude Code can execute without supervision. The scoping framework that makes parallel development work.
scopingworkflowclaude-code - Reviewing AI-Generated Code: The 3-Pass Method
AI writes confident-looking code that might be completely wrong. Here's how I review diffs across 6+ projects without letting bad code slip through.
code-reviewworkflowquality - Why Vibe Coding Breaks at Scale (And What Replaces It)
Vibe coding works for single tasks. It falls apart the moment you try to parallelize. Here's why, and the methodology that replaced it.
vibe-codingspec-codingmethodology - The Trust-But-Verify Approach to AI Code
The mental model for working with AI coding assistants without micromanaging or blindly accepting output. Trust the process, verify the diff.
methodologycode-reviewtrust - My Daily Workflow: Reviewing Diffs Across 6 Projects
A walkthrough of an actual Tuesday – scoping tasks, running agents, reviewing output, and shipping code across Scouter, Triumfit, Lucid, and more.
workflowdailydiff-review - Claude Code Project Setup for Multi-Repo Work
How to set up Claude Code for working across multiple repositories. CLAUDE.md per repo, terminal layout, context separation, and the git workflow that keeps everything reviewable.
claude-codesetupmulti-project - Solo Developer, Multiple Projects: How I Ship Progress on 6+ at Once
Context switching kills solo developers. Unless you change what context switching means. Here's how I ship across Scouter, Triumfit, Lucid, and more — every day.
solo-developerproductivityparallel - How to Write Specs That AI Can Execute
The tactical guide to writing specifications for Claude Code. The five elements, common mistakes, real before/after examples, and the specificity spectrum.
specstutorialclaude-code - The Parallel Development Playbook
The complete framework for running multiple AI coding sessions in parallel. Why parallel beats serial, the scoping requirement, review cadence, and where the diminishing returns hit.
parallelmethodologyworkflow - Vibe Coding vs Spec Coding
A direct comparison of vibe coding and spec coding. When each approach works, when each breaks down, the real cost of both, and the hybrid approach most people should be using.
vibe-codingspec-codingcomparison - 10 Common Claude Code Mistakes (and How to Fix Them)
The mistakes I see most often when developers use Claude Code for real projects. Each one has a fix that takes less than 5 minutes.
claude-codemistakestips - When Vibe Coding Is Actually Fine
Spec coding is my default. But vibe coding has legitimate use cases. Here's when conversational AI coding is the right call, and the anti-pattern of over-speccing trivial work.
vibe-codingmethodologyopinion - What Spec Coding Learned from Screenwriting
How writing four screenplays taught me to write better specifications for AI agents. The parallel between scripts and specs, and why cross-domain thinking produces better technical documents.
spec-codingscreenwritingmethodology - The Context Window Problem
Why long conversations with AI coding assistants degrade, and how self-contained specs solve the problem. The technical explanation behind the 'fresh session' approach.
claude-codecontext-windowspecs - The Solo Builder's Guide to AI Shipping
A complete guide for solo developers who want to ship more with AI coding assistants. The mindset shift, the daily rhythm, the toolchain, and the weekly output reality.
solo-developershippingguide - Building a SaaS with Claude Code
A real case study from building Scouter — a B2B lead enrichment SaaS. The spec phase, the build phase, the iteration phase, and what I learned about using AI to ship a production product.
case-studysaasclaude-code - From One Project to Eight: Scaling AI-Assisted Development
The progression from using AI on a single project to running 8 in parallel. What changes at each stage, what stays the same, and where the ceiling is.
scalingworkflowparallel - How to Delegate Coding Tasks to AI
A framework for deciding what to hand off to AI agents and how. The delegation spectrum, what makes a task delegatable, and the review contract you should never skip.
delegationai-agentsworkflow - Background Agents Explained
What background agents are, how to use them, and how to scale from 3 to 8 running simultaneously. The workflow, the prerequisites, and the failure modes.
background-agentsclaude-codeworkflow - Why Your AI Keeps Going Off Track
A diagnosis guide for when AI coding output misses the mark. The 5 reasons it happens, how to identify which one, and the spec fixes that correct each.
debuggingai-agentsspecs - Claude Code vs Cursor vs Copilot
An honest comparison from someone who's used all three on real projects. What each tool is best at, when to use each, and why the terminal won for parallel work.
comparisontoolsclaude-code