Does this skill work with existing projects?

Yes, it is designed to be initialized in any working directory to help Claude navigate and modify existing codebases with a clear roadmap.

Why should I use file-based planning with Claude?

It provides a 'working memory on disk,' allowing Claude to handle tasks that exceed its internal context window by persisting goals, findings, and progress in your local filesystem.

Where are the planning files stored?

Files like task_plan.md and findings.md are created directly in your current project root, ensuring they stay alongside your code and are easily accessible to you.

When is the best time to activate this skill?

You should use this skill for any complex task requiring more than 5 tool calls, research-heavy projects, or any multi-step workflow where organization is critical.

How does the 3-strike error protocol work?

It forces Claude to change its approach after a failure: first by diagnosing, second by trying an alternative method, and third by performing a broad rethink before escalating to the user.

Manus File-Based Planning

Name: Manus File-Based Planning
Author: jukasdrj

byjukasdrj

•

生産性とワークフロー

Implements a persistent markdown-based working memory for Claude to manage complex, multi-step coding and research tasks.

概要

The Planning with Files skill provides a structured framework for Claude to handle high-complexity tasks that exceed standard context limits. By maintaining persistent documentation like task_plan.md, findings.md, and progress.md directly in your project directory, Claude can 'remember' its state across dozens of tool calls. This skill enforces rigorous error-tracking and a 3-strike protocol to prevent repetitive failures, ensuring that long-running sessions remain organized, goal-oriented, and transparent to the user.

主な機能

Context-refresh hooks that read project plans before major actions
Continuous discovery logging to prevent multimodal data loss
Automated task plan initialization and phase tracking
Persistent 'Working Memory' via markdown files on disk
2 GitHub stars
3-Strike Error Protocol for intelligent failure recovery

ユースケース

Building complex new project modules from scratch with multi-file dependencies
Large-scale architectural refactoring requiring many interdependent steps
Deep research projects involving extensive web searching and data synthesis

概要

主な機能

Context-refresh hooks that read project plans before major actions
Continuous discovery logging to prevent multimodal data loss
Automated task plan initialization and phase tracking
Persistent 'Working Memory' via markdown files on disk
2 GitHub stars
3-Strike Error Protocol for intelligent failure recovery

ユースケース

Building complex new project modules from scratch with multi-file dependencies
Large-scale architectural refactoring requiring many interdependent steps
Deep research projects involving extensive web searching and data synthesis