Skill

Generate Actionable Project Plans

Writes short, specific, verifiable task plans - max 10 tasks per plan, no fixed templates, saved in the project root.


81
Spark score
out of 100
Updated last month
Version 13.1.0

Add to Favorites

Why it matters

Break down complex projects into small, verifiable tasks. This skill ensures clear goals, logical ordering, and specific, actionable steps for efficient project execution.

Outcomes

What it gets done

01

Define clear, concise tasks (2-5 minutes each).

02

Establish independent verification criteria for each task.

03

Structure tasks logically with identified dependencies.

04

Generate project plans tailored to specific project types (new, feature, bug fix).

Install

Add it to your toolbox

Run in your project directory:

curl -fsSL https://spark.entire.vc/get/ag-plan-writing | bash

Overview

Plan Writing

Provides a framework for writing short, specific, verifiable task plans without fixed templates, saved in the project root. Use when starting a new project, adding a feature, fixing a complex bug, or refactoring multiple files.

What it does

Provides a framework for breaking down work into clear, actionable tasks with verification criteria - emphasizing short, project-specific plans over rigid templates.

When to use - and when NOT to

Use this skill when starting a new project from scratch, adding a feature, fixing a complex bug, or refactoring multiple files. Not a fit for trivial single-step changes that don't need a written plan, and never use a fixed template copy-pasted across unrelated tasks - each plan must be unique to its specific task.

Inputs and outputs

Core task breakdown principles: each task should take 2-5 minutes with one clear outcome and be independently verifiable; verification must answer how you know it's done, what you can check or test, and what output is expected; ordering must identify dependencies, allow parallel work where possible, and always place a verification phase last; and plan files are saved as {task-slug}.md in the project root (e.g. "add auth" becomes auth-feature.md), never inside .claude/, docs/, or temp folders.

Five planning principles govern content: keep it short (5-10 clear tasks max, one page limit, no micro-step listing); be specific not generic (e.g. "Run npx create-next-app" instead of "Set up project", "Install next-auth, create /api/auth/[...nextauth].ts" instead of "Add authentication"); tailor content dynamically by project type (new project needs tech stack/MVP/file structure decisions, feature addition needs affected files/dependencies/verification, bug fix needs root cause/file-line/test); choose scripts by project type rather than copy-pasting all available scripts (frontend gets ux_audit.py/accessibility_checker.py, backend gets api_validator.py/security_scan.py, mobile gets mobile_audit.py, database gets schema_validator.py); and keep verification simple and concrete (e.g. "curl localhost:3000/api/users returns 200" instead of "Test the API").

The flexible plan structure is minimal: a one-sentence goal, a task checklist where each item pairs a specific action with a verification step, a "Done When" success criteria list, and optional notes - with no forced phases or sub-sections unless truly needed.

Best practices summarize the approach: start with the goal, cap plans at roughly 10 tasks (splitting into multiple plans if more are needed), ensure every task has clear done criteria, keep content project-specific rather than templated, and update checkboxes as work progresses.

Integrations

References project-type-specific validation scripts (ux_audit.py, accessibility_checker.py, api_validator.py, security_scan.py, mobile_audit.py, schema_validator.py) selected based on what the task actually touches.

Who it's for

Developers and agents writing implementation plans for new projects, features, bug fixes, or refactors who need short, concretely verifiable task lists rather than templated, over-specified project plans.

### Tasks
- [ ] Task 1: [Specific action] -> Verify: [How to check]
Source README

This skill provides a framework for breaking down work into clear, actionable tasks with verification criteria.

FAQ

Common questions

Discussion

Questions & comments · 0

Sign In Sign in to leave a comment.