TE

Scans the codebase for technical debt and generates trackable GitHub issues to manage code quality.

Install

mkdir -p .claude/skills/tech-debt && curl -L -o skill.zip "https://agentskills.codes/api/skills/download/5393" && unzip -o skill.zip -d .claude/skills/tech-debt && rm skill.zip

Installs to .claude/skills/tech-debt

Activation

This is the description your AI agent reads to decide when to run this skill — the better it matches your request, the more reliably it fires.

Technical debt management - scan codebase for bad smells and create tracking issues
83 charsno explicit “when” trigger
Beginner

Key capabilities

  • →Finds files exceeding 1000 lines
  • →Detects ESLint/oxlint suppression comments
  • →Identifies unsafe TypeScript 'any' type usages
  • →Locates risky mock patterns in test files
  • →Flags improper timer usage in tests
  • →Generates structured GitHub issues for debt remediation

How it works

It runs targeted shell commands (find, grep, awk) to pattern-match anti-patterns and code smells within the specified directory structure.

Inputs & outputs

You give it
Operation type (research or issue)
You get back
Technical debt report or created GitHub issue

When to use tech-debt

  • →Scanning for large files over 1000 lines
  • →Finding suppressed linting errors
  • →Identifying risky code patterns
  • →Creating tracking issues for technical debt

About this skill

Technical Debt

Operations

  • research [scope]: inspect the requested scope, defaulting to turbo/, and report verified findings. This operation is read-only.
  • issue [report]: create an English GitHub issue from current, verified findings. If no report exists, research first. Create or comment on an issue only when the caller requested that operation.

The tech-debt-research and tech-debt-issue commands remain aliases for these operations. A report is evidence at its recorded revision, not a standing instruction to change code.

Research

  1. Record the repository, HEAD, scope, and date. Read code quality and use the index for the affected surfaces.
  2. Use rg to find candidates and inspect matching files and consumers. Exclude dependencies, generated/vendor content, and historical migrations from generic cleanup recommendations. Preserve permanent migration records.
  3. Check type/lint suppressions, dynamic imports, environment configuration, error handling, fallback ownership, and unused dependencies against actual contracts. Use testing guidance for tests and ccstate guidance for signal/React code.
  4. Validate each candidate. File length, a relative import, a catch, or an ESLint off / Oxlint allow setting is not proof of a defect. Read override scope, replacement enforcement, generated-code boundaries, and documented exceptions. In particular, do not replace every floating promise with detach() or assume every parentless resetSignal() leaks.
  5. Report the concrete consequence, file/line, applicable rule, evidence, proposed remedy, and uncertainty. Use severity based on demonstrated impact. Distinguish static inspection from executable or production verification.

Keep the report proportional to the findings: scope and revision, confirmed issues, dismissed candidates when useful, and next actions. Save detailed evidence when it is too long for the response. Do not invent schedules or effort estimates from match counts.

Issue

Recheck the report against current source and existing issues before posting. Use an English title describing the concrete problem; include affected paths, evidence, impact, and actionable acceptance criteria in the body. Use only existing relevant labels and include source links pinned to the inspected SHA.

Write the body to a temporary file and pass gh issue create --repo okou-ai/okou --body-file <file> the exact Markdown. Keep the report focused enough to fit one issue; add detailed comments only when needed for the requested tracking task. Verify the created issue and return its URL. If posting fails, retain the report and state which action failed.

When not to use it

  • →Projects not using TypeScript or ESLint
  • →Quick prototyping phases where code quality is intentionally deprioritized
  • →Filesystems lacking a 'turbo' structure

Limitations

  • →Dependent on specific directory naming conventions
  • →May generate noise from intentional lint suppressions
  • →Limited to the predefined set of search patterns

How it compares

It systematically codifies code quality standards rather than relying on subjective ad-hoc reviews.

Compared to similar skills

tech-debt side by side with the closest alternatives in the catalog.

SkillInstallsUpdatedSafetyDifficulty
tech-debt (this skill)13moReviewBeginner
reviewing-nextjs-16-patterns1110moReviewIntermediate
dependency-upgrade06moReviewAdvanced
fix-dependabot-alerts188moReviewIntermediate

Try saying

Example prompts that trigger this skill in your AI assistant.

You might also like

reviewing-nextjs-16-patterns

djankies

Review code for Next.js 16 compliance - security patterns, caching, breaking changes. Use when reviewing Next.js code, preparing for migration, or auditing for violations.

11106

dependency-upgrade

pinkpixel-dev

Master major dependency version upgrades, compatibility analysis, staged upgrade strategies, and comprehensive testing approaches.

00

fix-dependabot-alerts

microsoft

Fix Dependabot security alerts by updating vulnerable npm dependencies. Use when the user mentions "dependabot", "security alerts", "vulnerability", "CVE", or wants to update packages with security issues.

1872

tech-debt-analyzer

ailabs-393

This skill should be used when analyzing technical debt in a codebase, documenting code quality issues, creating technical debt registers, or assessing code maintainability. Use this for identifying code smells, architectural issues, dependency problems, missing documentation, security vulnerabilities, and creating comprehensive technical debt documentation.

522

codex-code-review

tyrchen

Perform comprehensive code reviews using OpenAI Codex CLI. This skill should be used when users request code reviews, want to analyze diffs/PRs, need security audits, performance analysis, or want automated code quality feedback. Supports reviewing staged changes, specific files, entire directories, or git diffs.

16

code-audit

mei28

Automated code review tool that analyzes code quality, detects bugs, identifies security vulnerabilities, and suggests improvements based on industry best practices

00

Search skills

Search the agent skills registry