How It Works
The temporal evaluation pipeline
Every file in your repository has a complexity trajectory. GitVoyant extracts it, measures it, and classifies it.
Commit History
→
AST Analysis
→
Complexity Time Series
→
Linear Regression
→
Trend Classification
01
Extract
Retrieve commit history for each source file. Compute cyclomatic complexity at every snapshot using language-specific parsers.
02
Analyze
Fit linear regression to the complexity time series. Compute trend slope as complexity change per month. Score confidence based on commit history depth.
03
Classify
Classify each file as IMPROVING, DECLINING, or STABLE. Forecast quality decay risk bounded between 0 and 1. Surface the files that need attention.
gitvoyant analyze temporal ./project
| File | Trend | Score | Confidence | Risk |
|---|---|---|---|---|
| src/api/handlers.py | [-] DECLINING | +2.3 | 0.91 | HIGH |
| src/core/engine.go | [~] STABLE | +0.1 | 0.87 | LOW |
| src/utils/transform.js | [+] IMPROVING | -1.4 | 0.93 | LOW |
| src/models/User.java | [-] DECLINING | +1.8 | 0.85 | MEDIUM |
| src/services/auth.py | [+] IMPROVING | -0.6 | 0.90 | LOW |
| src/routes/index.ts | [~] STABLE | +0.0 | 0.72 | LOW |
Repository Health: STABLE | 6 files analyzed | 4 languages detected