plan-and-scratchpad

Automate creation, maintenance, and enforcement of PLAN.md and scratchpad.md artifacts.

Updated Feb 7, 2026
One-click install
npx skills add https://github.com/loagma/LoagmaPMS --skill plan-and-scratchpad
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: plan-and-scratchpad
Source: https://github.com/loagma/LoagmaPMS/tree/main/.claude/skills/plan-and-scratchpad
Command: npx skills add https://github.com/loagma/LoagmaPMS --skill plan-and-scratchpad

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Plan-and-scratchpad automates the creation, maintenance, and enforcement of PLAN.md and scratchpad.md artifacts to preserve context across agent turns.

Core Features & Use Cases

  • Establishes a formal contract between planning and execution phases with per-task PLAN.md and scratchpad.md.
  • Enforces worktree-scoped artifact paths to prevent parallel-agent context collisions and enable safe multi-task work.
  • Guides artifact lifecycle, phase discipline, and quality checks for onboarding new agents and complex projects.

Quick Start

Create or update PLAN.md and scratchpad.md for a new task following the defined structure and rules.

Frequently Asked Questions about plan-and-scratchpad

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I maintain agent context across turns during task execution?

You can maintain agent context across turns by automating the creation and enforcement of PLAN.md and scratchpad.md files, which preserve planning states and execution progress for continuous autonomous workflows.

What is the best way to prevent context collisions when running parallel agents?

Preventing context collisions in parallel agents requires enforcing worktree-scoped artifact paths, ensuring each agent operates within an isolated context boundary to safely manage multi-task work without overwriting shared state.

How do I onboard a new agent to an existing complex project workflow?

Onboarding a new agent involves applying structured artifact lifecycle rules and phase discipline, guiding the agent through existing PLAN.md and scratchpad.md contexts to quickly integrate into complex projects.

How does plan.md enforce phase discipline between planning and execution?

PLAN.md enforces phase discipline by establishing a formal contract between planning and execution phases, applying specific rules and anti-patterns to separate task planning from autonomous execution cleanly.

Do I need a specific framework to structure planning artifacts for agents?

No specific framework is required to structure planning artifacts for agents. The approach relies on standard markdown files and worktree-scoped paths to codify patterns and manage context persistence directly within your environment.