planning-with-files-de

What problem does it solve?

Complex, multi-step tasks that span multiple tool calls or sessions often lose critical context and progress when the AI's volatile working memory resets, leading to repeated work, missed requirements, and forgotten decisions.

Core Features & Use Cases

• Persistent File-Based Memory: Creates and maintains three core markdown files (task_plan.md for phase tracking, findings.md for research and technical decisions, progress.md for session logs) to store all task context permanently outside the AI's context window.
• Automatic Session Recovery: Detects unsynced progress from previous sessions after context clears (e.g., /clear command) and provides a structured catchup report to resume work without losing state.
• Structured Workflow Guardrails: Enforces rules like the 2-action rule (save findings after every 2 view/browser/search operations), mandatory error logging, and pre-decision plan reviews to reduce mistakes and repeated work.
• Use Case Example: Ideal for multi-step project planning, in-depth research tasks, complex bug diagnosis, and any work requiring organization across more than 5 tool calls, such as building a new software feature from scratch or conducting a full technical architecture review.

Quick Start

Use this skill to break down your upcoming multi-step project into tracked phases, capture all research findings, and automatically restore progress if your session resets.