planning-with-files

Maintain task_plan.md, findings.md, and progress.md for multi-step tasks.

Updated Apr 10, 2026
One-click install
npx skills add https://github.com/LuizEduPP/skills --skill planning-with-files-luizedupp
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: planning-with-files
Source: https://github.com/LuizEduPP/skills/tree/main/planning-with-files
Command: npx skills add https://github.com/LuizEduPP/skills --skill planning-with-files-luizedupp

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires orjson, and includes scripts (resource) components.

What problem does it solve?

This skill prevents complex, multi-step tasks from derailing when the AI loses context by forcing every major decision and discovery to live on disk, ensuring progress is retrievable even after session resets.

Core Features & Use Cases

  • Manus-style file set: task_plan.md, findings.md, and progress.md capture goals, discoveries, and session logs so no phase or insight is ever forgotten.
  • Automation hooks: Read, PreToolUse, PostToolUse, and Stop hooks keep plans fresh, require immediate logging, and invoke completion checks whenever sessions end.
  • Session recovery: Templates and scripts such as init-session and session-catchup streamline onboarding and aftermath analysis, making this ideal for research, project design, or debugging flows that span dozens of tool calls.

Quick Start

Use planning-with-files to scaffold task_plan.md, findings.md, and progress.md before starting any project that will take more than five tool calls.

Frequently Asked Questions about planning-with-files

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

FAQPage Schema
How do I maintain task state and recover progress when AI loses context during long sessions?

To maintain task state during long sessions, use file-based planning to persist goals, discoveries, and logs on disk. This ensures progress is retrievable and complex multi-step tasks resume smoothly even after session resets.

What is the best way to manage complex multi-step development tasks that require persistent state?

The best way to manage complex multi-step tasks is using a Manus-style file set like task_plan.md, findings.md, and progress.md. These files capture goals and discoveries, preventing project derailment when context is lost.

How do I scaffold a task plan and track findings for research projects spanning dozens of tool calls?

Scaffold research projects by initializing task_plan.md, findings.md, and progress.md files before starting. Helper scripts automate initialization and completion checks, streamlining onboarding and aftermath analysis for extensive tool flows.

Do I need automation hooks to keep task plans fresh across multiple sessions?

Yes, automation hooks like PreToolUse and PostToolUse keep task plans fresh by requiring immediate logging. Stop hooks invoke completion checks whenever sessions end, ensuring all major decisions live on disk.

When should I use file-based planning for project work instead of relying on session memory?

Use file-based planning whenever research, development, or project work requires more than five tool calls or needs persistent state across sessions. It forces every major decision onto disk, preventing complex tasks from derailing.

Does planning-with-files work with orjson for managing session recovery scripts?

Yes, planning-with-files depends on orjson and uses helper scripts for session catch-up and initialization. These scripts manage the task_plan.md and progress.md files to ensure state is recovered after interruptions.