Create Plans

Turns project briefs into executable plans optimized for Claude as the builder, not enterprise documentation. The core insight is smart: plans are prompts, not documents that get converted to prompts later. It enforces aggressive scope control because Claude's quality degrades around 40-50% context usage when it sees work piling up, not at 80% like you'd expect. Plans split into 2-3 task chunks with built-in deviation rules so bugs get auto-fixed without breaking flow. Heavy emphasis on CLI automation over manual checkpoints, milestone-driven shipping, and treating git history as your project timeline. Opinionated against team coordination overhead since you're planning for one person and one AI.

skills.sh installs: 200
GitHub stars: 2K
last updated: Aug 21, 2026

View on GitHub

Install

$ npx -y skills add glittercowboy/taches-cc-resources --skill create-plans --agent claude-code

Installs into .claude/skills of the current project.

Files

<essential_principles>

You are planning for ONE person (the user) and ONE implementer (Claude). No teams. No stakeholders. No ceremonies. No coordination overhead. The user is the visionary/product owner. Claude is the builder. PLAN.md is not a document that gets transformed into a prompt. PLAN.md IS the prompt. It contains: - Objective (what and why) - Context (@file references) - Tasks (type, files, action, verify, done, checkpoints) - Verification (overall checks) - Success criteria (measurable) - Output (SUMMARY.md specification)

When planning a phase, you are writing the prompt that will execute it.

Plans must complete within ~50% of context usage to maintain consistent quality.

The quality degradation curve: