TD Agent

Public

An AI technical-debt analyser built as an MSc dissertation, benchmarked directly against SonarQube.

Python

FastAPI

Next.js

Gemini

Python

View CodeDemo coming soon

// architecture

LLM-driven commit history analysis · SQALE-matched scoring · git churn/hotspot detection · background jobs + live dashboard

// overview

Analyses a repository's full commit history with an LLM and scores technical debt in remediation minutes, deliberately unit-matched to SonarQube's SQALE index so the two are directly comparable. Includes git churn and hotspot analysis, smart file prioritisation within token budgets, and background analysis jobs feeding a live dashboard.

·Analyses a repository's full commit history with an LLM, not just a single snapshot.
·Scores technical debt in remediation minutes, deliberately unit-matched to SonarQube's SQALE index for direct comparison.
·Git churn and hotspot analysis surfaces files that change often and carry the most risk.
·Smart file prioritisation keeps large repos within LLM token budgets.
·Background analysis jobs feed a live dashboard rather than blocking on a single request.
·Early results show the LLM surfacing architectural debt that static analysis misses.

Architecture diagram

coming soon

Demo walkthrough

coming soon