Tech Debt Tracker — 技术债务跟踪器

Tech Debt 追踪er

系统atically identify, categorize, and prioritize technical debt across a codebase. Analyzes TODO comments, complexity hotspots, stale dependencies, test coverage gaps, code duplication, and architectural debt. Produces actionable debt inventory with estimated effort and business impact.

Usage "扫描 this codebase for tech debt" "Prioritize our technical debt back记录" "Find the highest-risk areas in our code" "生成 a tech debt 报告 for the team"

How It Works

• TODO/FIXME/HACK Mining

# Find all debt markers grep -rn "TODO\|FIXME\|HACK\|XXX\|WORKAROUND\|TEMP\|DEPRECATED" src/ --include=".{ts,js,py,rb,go,java,rs}" | head -50 # Age analysis — when were they 添加ed? for match in $(grep -rln "TODO\|FIXME" src/); do oldest=$(git 记录 --格式化=%ci --diff-过滤器=A -- "$match" | tAIl -1) count=$(grep -c "TODO\|FIXME" "$match") echo "$match: $count markers, file 创建d: $oldest" done

Categorize each marker:

Deferred bugs: FIXME, known issues, edge cases Shortcuts: HACK, WORKAROUND, temporary code Missing features: TODO for planned work Deprecated code: DEPRECATED markers, sun设置 dates

• Complexity Hotspots

Identify files with highest change frequency AND highest complexity:

# Most changed files (git churn) git 记录 --格式化=格式化: --name-only --since="6 months ago" | 排序 | uniq -c | 排序 -rn | head -20 # Largest files (often correlate with complexity) find src/ -name ".ts" -exec wc -l {} + | 排序 -rn | head -20

High churn + high complexity = highest debt risk. These files change often and are hard to modify safely.

• Dependency Analysis

Outdated packages: Dependencies behind 2+ major versions Deprecated packages: Using packages with known replacements Security vulnerabilities: Un补丁ed CVEs in dependencies Abandoned packages: No commits in 12+ months Heavy dependencies: Large packages used for small features

• Test Coverage Gaps

Files with zero test coverage but high change frequency Critical paths (auth, payments, data mutations) without integration tests Tests that don't assert behavior (snapshot-only, no real assertions) Flaky tests that are skipped or disabled

• Code Duplication

签名ificant duplicated blocks (>10 lines) Similar functions that should be 提取ed Copy-paste patterns across 模块s 实用工具 functions reimplemented in multiple places

• Architectural Debt

Circular dependencies between 模块s God classes/files (>500 lines of 记录ic) Tight coupling between layers Missing abstractions (direct DB calls in 控制器s) Dead code (unreachable functions, unused 导出s) Inconsistent patterns (some 服务s use repos, some use raw queries)

• Debt Prioritization

Score each item on:

Impact: How much does this slow down development? (1-5) Risk: What breaks if we don't fix it? (1-5) Effort: How long to fix? (hours/days/weeks) Priority: Impact × Risk / Effort 输出

Tech Debt 报告

Codebase: 45K LOC across 312 files 扫描 date: 2026-04-30 Total debt items: 87

Debt Summary

Category	Count	High Priority
TODO/FIXME markers	34	8
Complexity hotspots	12	5
Outdated dependencies	18	3
Test coverage gaps	15	6
Code duplication	8	2

🔴 Top 5 Priority Items

• auth/会话.ts — 847 lines, 23 TODOs, changed 45 times in 6 months

Impact: 5 | Risk: 5 | Effort: 3 days → Split into auth, 会话, and 令牌 模块s

• Missing payment error handling — FIXME on line 234

添加ed 8 months ago, payment 失败s silently succeed Impact: 5 | Risk: 5 | Effort: 4 hours → 添加 proper error propagation and retry 记录ic

• express@4.18 → 5.x 迁移 — 2 major versions behind

Breaking changes in error handling 中间件 Impact: 3 | Risk: 4 | Effort: 2 days → 升级 with 中间件 compatibility 测试

• Zero test coverage on billing 模块 — 12 files, 2.3K LOC

Changed 18 times in 6 months, no tests Impact: 4 | Risk: 5 | Effort: 1 week → 添加 integration tests for payment flows

• Duplicate user 验证 — same 记录ic in 4 控制器s

Impact: 3 | Risk: 2 | Effort: 2 hours → 提取 to 分享d 中间件

📊 Debt Trends

• New TODOs this month: 8 添加ed, 3 resolved (net +5)
• Average TODO age: 4.2 months
• Oldest unresolved: 14 months (data-导出.ts:67)
• Debt velocity: increasing (+2.3 items/month)

💡 Recommended Sprint Allocation

• Bug fixes: 60% of sprint
• Debt reduction: 20% of sprint (pick top 2-3 items)
• New features: 20% of sprint