thoroughness-scoring

Scores implementation options with a 1-10 thoroughness rating and recommends the highest-quality choice.

2.8k|277|Updated Feb 1, 2026
One-click install
npx skills add https://github.com/rohitg00/pro-workflow --skill thoroughness-scoring
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: thoroughness-scoring
Source: https://github.com/rohitg00/pro-workflow/tree/main/skills/thoroughness-scoring
Command: npx skills add https://github.com/rohitg00/pro-workflow --skill thoroughness-scoring

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

When AI assistance makes thorough work nearly as fast as quick hacks, developers still default to shortcuts that accumulate technical debt. This Skill forces every decision point through a structured scoring rubric so the higher-quality option is chosen deliberately, not by habit.

Core Features & Use Cases

  • Thoroughness Rating Scale: Scores every option T:1-10 based on edge cases, tests, types, docs, and error handling, with a defined rubric for each level.
  • Effort Delta Comparison: Presents manual vs AI-assisted time estimates side by side so the marginal cost of doing things properly is explicit.
  • Scope Classification: Distinguishes contained work (score and execute) from unbounded work (break into shippable pieces before scoring).
  • Use Case: When adding rate limiting to an API endpoint, the Skill presents a full sliding-window Redis-backed limiter (T:9, ~20 min with AI) versus a basic in-memory counter (T:4, ~5 min), then recommends the thorough option with the concrete delta.

Quick Start

Ask the AI to score the implementation options for your current task using thoroughness scoring and recommend the highest-rated approach.

Frequently Asked Questions about thoroughness-scoring

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

FAQPage Schema
How do I choose between a quick hack and a thorough implementation?

Score each option on a T:1-10 thoroughness scale covering edge cases, tests, types, and docs, then compare manual versus AI-assisted effort. When AI closes the time gap, recommend the T:8+ option and state the concrete delta it adds.

What is a thoroughness rating scale for code decisions?

It is a 1-10 rubric where T:10 means full edge cases, tests, docs, and helpful errors, while T:1 is an untested copy-paste hack. Each option presented to the user gets a score plus effort estimates with and without AI assistance.

When should I skip thoroughness scoring?

Skip scoring for exploratory spikes, one-off throwaway scripts, debugging sessions, and learning exercises where iteration speed matters more than completeness. State explicitly that scoring was skipped so the choice is conscious.

How do I handle tasks that are too large to do thoroughly?

Classify the scope first: unbounded work like refactoring all error handling must be broken into contained, independently shippable pieces. Score each piece individually rather than attempting T:10 on the whole unbounded scope.

What are common mistakes when scoring implementation options?

Anti-patterns include scoring all options identically, using low scores to justify shortcuts, omitting the effort comparison, and treating T:10 as the default target on unbounded work. Always re-state project, branch, and task context before presenting options.