What is the benefit of file-based planning for Claude?

It acts as an external 'hard drive' for the AI's memory, ensuring that goals and findings aren't lost when the conversation becomes long or context-heavy.

How does the 3-strike error protocol work?

It forces the AI to first diagnose and fix, then try an alternative approach, and finally rethink the entire strategy before escalating to the user, preventing repetitive failures.

When should I use the planning-with-files skill?

It is recommended for any complex task involving more than 5 tool calls, such as building new features, deep research, or multi-file refactoring.

Where are the planning files stored?

While templates reside in the skill directory, your active planning files (task_plan.md, findings.md, and progress.md) are created directly in your project's root directory.

Manus-Style Task Planning

Name: Manus-Style Task Planning
Author: iseepatterns-jz

byiseepatterns-jz

生产力与工作流

Implements a persistent file-based workflow to manage context and state during complex multi-step coding and research tasks.

关于

Planning with Files introduces a robust, Manus-style memory management system for Claude Code by offloading complex project state to persistent Markdown files. By maintaining a task_plan.md, findings.md, and progress.md directly in the project directory, this skill ensures that critical goals, research data, and error logs remain accessible even as the AI's context window shifts. It is particularly effective for long-running research projects, multi-step development tasks, or any operation requiring more than five tool calls, enforcing a plan-first discipline and a structured three-strike error resolution protocol to prevent repetitive failures.

主要功能

Structured 3-strike error protocol for systematic debugging and escalation.
Persistent 'Working Memory on Disk' to prevent critical information loss during long sessions.
Automatic generation of task_plan.md, findings.md, and progress.md templates.
Pre-decision checkpoints that refresh goals within the AI's attention window.
0 GitHub stars
The 2-Action Rule for immediate saving of browser and research findings.

使用场景

Debugging intricate, elusive bugs that require tracking multiple failed attempts and hypotheses.
Executing complex, multi-step software development projects from scratch.
Conducting deep research tasks that require spanning multiple browser and tool sessions.

关于

主要功能

Structured 3-strike error protocol for systematic debugging and escalation.
Persistent 'Working Memory on Disk' to prevent critical information loss during long sessions.
Automatic generation of task_plan.md, findings.md, and progress.md templates.
Pre-decision checkpoints that refresh goals within the AI's attention window.
0 GitHub stars
The 2-Action Rule for immediate saving of browser and research findings.

使用场景

Debugging intricate, elusive bugs that require tracking multiple failed attempts and hypotheses.
Executing complex, multi-step software development projects from scratch.
Conducting deep research tasks that require spanning multiple browser and tool sessions.