What is Manus-style planning?

It is a workflow pattern that treats the local filesystem as 'external RAM' for the AI, ensuring that goals, findings, and progress are persisted outside of the limited context window.

Does this skill help with session timeouts or clearing history?

Yes, the skill includes a session-catchup script that reads your git diffs and existing planning files to restore Claude's context immediately after a session reset.

When should I use this skill instead of standard Claude Code commands?

Use this skill for any task requiring more than 5 tool calls, research-heavy projects, or complex builds that span multiple development phases.

Where are the planning files stored?

While the skill templates live in the plugin root, your active planning files (task_plan.md, etc.) are stored directly in your project's root directory for persistence and visibility.

Manus-Style Planning with Files

Name: Manus-Style Planning with Files
Author: OthmanAdi

byOthmanAdi

•

10,422

Productivity & Workflow

Implements a persistent, file-based workflow to manage complex AI coding tasks and maintain long-term project context.

About

This skill provides Claude Code with a 'working memory on disk' by utilizing structured markdown files—task_plan.md, findings.md, and progress.md—to maintain context across long sessions. Inspired by the Manus workflow pattern, it enforces disciplined habits such as documenting errors, updating project state after tool calls, and performing automatic session recovery after context clears. It is specifically designed for developers tackling multi-step architectural changes or deep-dive research where Claude's standard context window might become saturated or volatile, ensuring the AI never repeats failures and always knows its current phase.

Key Features

Automatic session recovery and context re-syncing to handle /clear commands or new sessions.
Hook-based monitoring that triggers plan updates after file edits or tool usage.
The 2-Action Rule for research to ensure multimodal data is captured in text before it is lost.
10,422 GitHub stars
Persistent task tracking via task_plan.md, findings.md, and progress.md templates.
3-Strike Error Protocol to prevent repetitive AI loops and ensure approach mutation.

Use Cases

Conducting deep research tasks that involve extensive web searching and data synthesis.
Managing complex, multi-step code refactoring or large-scale feature implementations.
Resuming long-running AI-driven projects with full context restoration from previous sessions.

About

Key Features

Automatic session recovery and context re-syncing to handle /clear commands or new sessions.
Hook-based monitoring that triggers plan updates after file edits or tool usage.
The 2-Action Rule for research to ensure multimodal data is captured in text before it is lost.
10,422 GitHub stars
Persistent task tracking via task_plan.md, findings.md, and progress.md templates.
3-Strike Error Protocol to prevent repetitive AI loops and ensure approach mutation.

Use Cases

Conducting deep research tasks that involve extensive web searching and data synthesis.
Managing complex, multi-step code refactoring or large-scale feature implementations.
Resuming long-running AI-driven projects with full context restoration from previous sessions.