skills.rest
Herxinsasa
HerxinsasaCOMMUNITY
Explore all skills by this author →
GitHub

planning-with-files

Track multi-step AI agent tasks using persistent markdown files.

1|Updated Apr 22, 2026
Search Tags:#context management#agent workflow#progress tracking#task planning#session recovery#multi-step tasks#file-based planning
One-click install
npx skills add https://github.com/Herxinsasa/Skills-Collector --skill planning-with-files-herxinsasa
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: planning-with-files
Source: https://github.com/Herxinsasa/Skills-Collector/tree/main/planning-with-files/.cursor/skills/planning-with-files
Command: npx skills add https://github.com/Herxinsasa/Skills-Collector --skill planning-with-files-herxinsasa

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve?

This Skill eliminates the risk of losing track of goals, progress, and critical discoveries during long, multi-step tasks that require dozens of tool calls, where volatile AI context windows cause important information to be forgotten between interactions.

Core Features & Use Cases

  • Persistent External Memory: Creates and maintains three core markdown files (task_plan.md for phase tracking, findings.md for research and decisions, progress.md for session logs) that act as a durable working memory on disk.
  • Session Recovery: Automatically catches up on unsynced context from previous sessions after a context reset, so work can resume without losing state.
  • Structured Workflow Guardrails: Enforces rules like the 2-Action Rule to save multimodal findings immediately, 3-strike error protocols to avoid repeating failed actions, and mandatory plan re-reading before major decisions to keep goals in focus.
  • Use Cases: Ideal for complex project planning, bug diagnosis, research tasks, feature development, and any work spanning 5 or more tool calls that requires structured tracking.

Quick Start

Use this Skill to plan and track your complex multi-step task by having the AI create and update task_plan.md, findings.md, and progress.md files in your project directory as it works, with automatic recovery if you reset the context mid-task.

Frequently Asked Questions about planning-with-files

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

FAQPage Schema
How do I prevent context loss during long multi-step AI agent tasks?▼

Prevent context loss during long multi-step AI agent tasks by using persistent markdown files as external working memory to track phases, log findings, and record progress. This ensures critical discoveries survive context window resets.

How does file-based planning work for complex project planning and bug diagnosis?▼

File-based planning works by creating task_plan.md, findings.md, and progress.md on disk to structure phase-based tracking and mandatory findings logging. This enforces workflow rules like saving multimodal discoveries after every two view operations.

Can I recover my agent workflow state after a session reset?▼

You can recover agent workflow state after a session reset through automatic context catchup. The persistent markdown files synchronize unsynced context from previous sessions, allowing work to resume without losing state.

When do I need persistent external memory for agent workflows?▼

You need persistent external memory for agent workflows when executing complex tasks requiring five or more tool calls. It is essential for feature development, research tasks, and bug diagnosis that require structured progress tracking across multiple interactions.

What is the best way to track progress for multi-step tasks with volatile context windows?▼

The best way to track progress for multi-step tasks with volatile context windows is implementing phase-based task planning and error tracking with three-strike failure protocols. This approach maintains goal focus by enforcing mandatory plan re-reading before major decisions.

Does file-based task planning require any external dependencies?▼

File-based task planning does not require external dependencies. It relies solely on creating and maintaining markdown files within your project directory to serve as a durable working memory for the agent.

Related Skills

Explore More Software Engineering Skills→
jinglemansweepjinglemansweep

jp-setup

Create a .plans directory with a .gitkeep file at the repository root.

Community
Basic
i3ringiti3ringit
3

file-todos

Manage Markdown todo files with YAML frontmatter in the todos/ directory.

Community
Basic
OKHP3OKHP3
2

okhp3-foundry-repo-creator

Convert AI-driven assets and ideas into standardized GitHub repositories.

Community
Advanced
jonnymuirjonnymuir
1

nap

Compress agent history files and prune stale logs in the .squad/ directory.

Community
Basic