What’s the best way you’ve found to track technical debt? Tools, frameworks, metrics, etc.

Some of the best ways that have worked to track technical debt are metrics and graphs.

I use ClickUp, in it the card that will go to Tech Debt status it must have:

All comments regarding what happened to this card, thus to be considered tech debt.

Links to the documentation and repos.

The reasons we decided to become one, the expectations, reality, outcomes and the scenarios needed to leave this status, so the card could end In Progress.

I use Project Management and Issue Tracking Tools:
Integrate debt items directly into your workflow alongside other tasks.
personally I use  GitHub Issues, also you need to use Code Analysis Tools:
Detect and measure technical debt within the codebase.
Offer insights into code quality, maintainability, and potential issues. I prefer  SonarQube. 
Metrics to Measure Technical Debt:

Code Coverage: Percentage of code covered by automated tests.
Defect Ratio: Ratio of new bugs to closed bugs (indicates code quality).
Lead Time: Time between task commitment and release (longer lead times can suggest debt).
Technical Debt Ratio (TDR): Estimated effort to fix debt items divided by total codebase size.
Debt Index: Similar to TDR, calculated by dividing debt (in hours or story points) by codebase size (lines of code or function points).
The most effective approach will depend on your specific team, codebase, and organizational context. Experiment with different methods and tools to find what works best for you.

One great way for tracking tech debt is making it tracked as a "risk".