| name | tech-debt-tracker |
| description | Agents should invoke this skill when identifying, categorizing, prioritizing, or planning technical debt work, debt sprints, cleanup backlogs, TODO consolidation, or long-term maintainability risks. Tracks debt with severity/effort. |
Tech Debt Tracker
Systematic tracking and management of technical debt. Debt is not inherently bad — it's a trade-off. The problem is untracked debt that compounds silently.
Quick Start
Log a New Debt Item
When debt is identified (during reviews, development, or ad-hoc):
- Categorize: What type of debt is it?
- Severity: How much pain does it cause?
- Effort: How much work to fix?
- Priority: Severity vs effort — is it worth fixing now?
- Log: Add to the registry in
MEMORY.md
Debt Categories
| Category | Description | Examples |
|---|
| Architecture | Structural issues that affect the whole system | Missing module boundaries, circular dependencies, wrong abstraction level |
| Code | Local code quality issues | Code smells, duplicated logic, overly complex functions, dead code |
| Test | Missing or inadequate test coverage | No tests for critical paths, brittle tests, missing edge case coverage |
| Documentation | Missing or outdated documentation | Undocumented public APIs, stale README, missing architecture docs |
| Dependency | Outdated or problematic dependencies | Deprecated libraries, unpatched CVEs, abandoned packages, version conflicts |
Severity Levels
| Severity | Meaning | Impact |
|---|
| Critical | Blocks progress or causes recurring incidents | Developers work around it daily, causes bugs |
| High | Causes ongoing friction and slows development | Every feature touching this area takes longer |
| Medium | Noticeable but manageable | Annoyance, not a blocker. Occasional confusion. |
| Low | Cosmetic or minor | Cleanup opportunity, no functional impact |
Effort Estimates
| Effort | Time | Scope |
|---|
| S | < 1 hour | Single function or file change |
| M | 1-4 hours | Multiple files, focused refactoring |
| L | 4-16 hours | Cross-module changes, possible API updates |
| XL | > 16 hours | Major restructuring, migration, or rewrite |
Priority Matrix
Use severity + effort to determine action:
| Effort S | Effort M | Effort L | Effort XL |
|---|
| Critical | Fix immediately | Fix this sprint | Plan and schedule | Plan, break into phases |
| High | Fix now (quick win) | Fix this sprint | Schedule within 2 weeks | Plan, break into phases |
| Medium | Fix when nearby | Schedule within 2 weeks | Backlog | Backlog (consider living with it) |
| Low | Fix if touching the file | Backlog | Backlog | Accept the debt |
Quick wins (High severity + S effort, or Critical + S/M effort) should be fixed as soon as possible — they deliver the most value per hour of investment.
Debt Entry Format
When adding to the registry in MEMORY.md:
| Field | Description |
|---|
| ID | Incremental number (DEBT-001, DEBT-002, ...) |
| Project | Which project/codebase |
| Category | Architecture / Code / Test / Documentation / Dependency |
| Description | What the debt is and why it matters |
| Severity | Critical / High / Medium / Low |
| Effort | S / M / L / XL |
| Status | Open / In Progress / Resolved / Accepted |
| Discovered | Date the debt was identified |
| Resolved | Date it was fixed (if applicable) |
Example entry:
| ID | Project | Category | Description | Severity | Effort | Status | Discovered |
|---|
| DEBT-001 | tui-setup | Code | app.rs has cyclomatic complexity 32 in handle_event — needs decomposition | High | M | Open | 2026-02-14 |
| DEBT-002 | tui-setup | Test | No unit tests for parser module — refactoring is risky | High | L | Open | 2026-02-14 |
Debt Review Cadence
When to Review Debt
- During code reviews: Note new debt found during review
- Start of major feature work: Check if the area has known debt that should be addressed first
- Monthly review: Scan the registry for items that have changed priority (became more/less urgent)
- After incidents: Did the debt contribute? Escalate priority if so
Monthly Debt Review Checklist
- Scan the debt registry for stale items (resolved but not marked, or no longer relevant)
- Re-assess priority of open items (has context changed?)
- Identify quick wins that have accumulated
- Propose a "debt sprint" if high-severity items are piling up
- Archive resolved items (move to a "Resolved" section in MEMORY.md)
Debt Reporting
Debt Summary Report
When asked for a debt status report:
## Tech Debt Summary: [Project]
**Date:** YYYY-MM-DD
**Total items:** N (C critical, H high, M medium, L low)
### Quick Wins Available
[S-effort items with High+ severity — easy fixes with big impact]
| ID | Description | Effort |
|---|---|---|
| DEBT-003 | ... | S |
### Top Priority Items
[Highest severity items regardless of effort]
| ID | Description | Severity | Effort |
|---|---|---|---|
| DEBT-001 | ... | Critical | M |
### Trend
[Is debt growing, shrinking, or stable? Any concerning patterns?]
### Recommendations
1. [Prioritized actions]
Integration
- Data: Debt registry maintained in
MEMORY.md (Tech Debt Registry section)
- Cross-reference: architecture-review (discovers structural debt), code-quality (discovers code debt), refactoring-advisor (plans debt remediation)
- Security debt: If debt has security implications (unpatched dependencies, missing input validation), flag to Zero
Arc skill — Technical debt tracking and management
