How Cyclopt analyzes your code

Both Cyclopt Companion and Cyclopt Panorama are powered by the same analysis engine. This page explains what happens when Cyclopt analyzes a commit and how to read the results.

A data-driven, hierarchical methodology

Cyclopt follows a hierarchical quality-evaluation methodology aligned with the ISO/IEC 25010:2023 standard. It relies on measurements that quantify major source-code properties (complexity, coupling, cohesion, level of proper documentation, and compliance with widely accepted coding practices). Using this data, Cyclopt evaluates your code across four quality characteristics:

• Maintainability: how easy your code is to change safely.
• Security: how exposed your code is to vulnerabilities.
• Readability: how easy your code is to understand.
• Reusability: how easily parts of your code can be reused.

An overall quality score is derived from these characteristics.

Scores and letter grades

Each characteristic and the overall quality are expressed as a numeric score and a letter grade shown as a color-coded avatar. Grades run from best to worst:

A⁺ · A · A⁻ · B⁺ · B · B⁻ · C⁺ · C · C⁻ · D⁺ · D · D⁻

Alongside each grade you'll usually see a weekly trend arrow indicating whether that characteristic improved or declined since the previous week.

What Cyclopt measures

Every analysis produces a set of findings you can drill into:

Violations

Rule breaches detected in your source code, grouped by category and by severity: critical, major and minor. Every violation comes with an explanation. The full, per-language catalog of rules lives in the Violations reference.

Vulnerabilities

Security issues, including dependency vulnerabilities classified by severity (critical, high, moderate, low).

Duplication

Duplicate code across your repository: duplicate instances, duplicated lines of code, the classes that contain clones, and the overall duplicate-code percentage.

Coverage

Where test-coverage data is available, Cyclopt reports covered / partially covered / uncovered lines and per-metric coverage (statements, functions, branches, classes, methods, instructions).

Metrics

A detailed, per-metric breakdown (logical and physical lines of code, number of files / classes / functions, maintainability index, cyclomatic complexity, comment density, code effort, and more). Each color-coded and, where relevant, accompanied by a suggested fix.

Branches

Analysis is performed per branch. Cyclopt gives special meaning to a repository's production and staging branches, as these are the branches used by default in quality gates and early alerts, and they're the ones you'll most often compare.

Tuning what gets analyzed

You can exclude branches, files, folders and specific rules and tweak analysis parameters with a cyclopt.yaml file in your repository. See the Configuration file reference.

To silence individual findings directly in your source code, use inline cyclopt-ignore comments.