tech-debt

Scan, categorize, and prioritize technical debt in a codebase register.

13|Updated Jul 30, 2026
One-click install
npx skills add https://github.com/frabcd/codex-ai-game-studio --skill tech-debt-frabcd
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: tech-debt
Source: https://github.com/frabcd/codex-ai-game-studio/tree/main/plugins/ai-game-studio/skills/tech-debt
Command: npx skills add https://github.com/frabcd/codex-ai-game-studio --skill tech-debt-frabcd

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Technical debt accumulates silently through TODO comments, workarounds, and outdated dependencies, making it hard to know what to fix first. This Skill scans the codebase for debt indicators, maintains a structured debt register, and scores items so teams can schedule repayment deliberately. ## Core Features & Use Cases - Debt Scanning: Detects TODO, FIXME, HACK, and @deprecated markers, duplicated code, oversized files, and long functions, then categorizes findings into architecture, code quality, test, documentation, dependency, and performance debt. - Debt Register Management: Adds manual entries with category and effort estimates, and maintains a Markdown register at docs/tech-debt-register.md with user confirmation before every write. - Prioritization and Reporting: Scores items by impact, frequency, and fix effort, re-sorts the register, and generates trend reports flagging items older than three sprints. - Use Case: Before sprint planning, run a scan to catch new debt, re-prioritize the register, and feed the top items into the sprint plan. ## Quick Start Use the tech-debt skill to scan this codebase for debt indicators and write the findings to the debt register.

Frequently Asked Questions about tech-debt

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

FAQPage Schema
How do I scan a codebase for technical debt?

Run the scan subcommand to search for TODO, FIXME, HACK, and @deprecated markers, duplicated code blocks, files over 500 lines, and functions over 50 lines. Findings are categorized into six debt types and can be written to a register after your approval.

How do I prioritize technical debt items for a sprint?

Use the prioritize mode, which scores each register entry by impact multiplied by encounter frequency divided by fix effort. The register is re-sorted by score and the top items are recommended for the next sprint.

What categories of technical debt does the scan detect?

Findings are grouped into architecture debt, code quality debt, test debt, documentation debt, dependency debt, and performance debt. Each category maps to specific indicators like coupling issues, duplication, missing tests, or outdated packages.

Does the tech debt scan modify my files automatically?

No. Every write to docs/tech-debt-register.md requires explicit user confirmation, and report mode is fully read-only. If you decline a write, the workflow stops with a blocked verdict.

When should technical debt items be removed from the register?

Items older than three sprints without action should either be fixed or consciously accepted with a documented reason. Running a report at the start of each sprint surfaces these stale entries and tracks whether debt is growing or shrinking.