AtlasArc Architecture Report

Key Findings

HighCycles

1 problem cycle group reported in this report scope.

NeutralCoupling

NCCD is 2.79. Use it as a same-project dependency-shape baseline; inspect cycles and concrete paths before interpreting a change.

NeutralJVM public surface

ARV and GRV describe JVM public-surface breadth: ARV 0.62, GRV 0.51. Use package RV, marked public classes, real callers, and intended boundaries to interpret the result.

NeutralGovernance

No Safe Havens or repository governance records are active; all findings are treated as active structural findings.

Highest-Value Targets

spring-boot-build.core.spring-boot.main:org.springframework.bootArchitectural cycle participant, fan-in 169 / fan-out 8
spring-boot-build.core.spring-boot.main:org.springframework.boot.context.properties.bindArchitectural cycle participant, fan-in 44 / fan-out 3
spring-boot-build.core.spring-boot.main:org.springframework.boot.context.properties.sourceArchitectural cycle participant, fan-in 32 / fan-out 2, distance 0.67
spring-boot-build.core.spring-boot.main:org.springframework.boot.envArchitectural cycle participant, fan-in 11 / fan-out 2
spring-boot-build.core.spring-boot-autoconfigure.main:org.springframework.boot.autoconfigure.service.connectionfan-in 63 / fan-out 0, distance 0.60, visibility 0.71

The report reads package structure and dependency evidence; it does not know what the code is meant to do.
Coverage metrics are not loaded, so coverage-dependent risk readings are unavailable.
Package exclusions are disabled for this report context.

What This Report Says

Cycle risk is the dominant signal

Read this report as cycle-first evidence: resolve or consciously accept the reported cycle groups before treating lower-level metrics as optimization work.

1 problem cycle group reported in this report scope.
First package to inspect: spring-boot-build.core.spring-boot.main:org.springframework.boot (Architectural cycle participant, fan-in 169 / fan-out 8).
Largest cycle group: Problem cycle 0 with 4 packages.

Evidence Summary

NCCD2.79

Same-project dependency-shape baseline; compare only on an equivalent analysis basis.

ARV0.62

Unweighted average of package Relative Visibility; each package contributes equally.

GRV0.51

Public JVM classes divided by all JVM classes across the graph.

Cycle groups1

Cycle clusters remain in the report scope.

Precise graph statistics remain available below so the report stays auditable.

CCD	Cumulative Component Dependency	25751.00
ACD	Average Component Dependency	25.15
RACD	Relative Average Component Dependency	0.02
NCCD	Normalized Cumulative Component Dependency	2.79
ARV	Average Relative Visibility	0.62
GRV	Global Relative Visibility	0.51

Architecture Overview Focus

This profile reads the whole model first: cycle posture, highest inspection target, and any governance context that changes how warnings should be interpreted.

Suggested Next Steps

1Start with Problem cycle 0

4 packages participate in this architectural cycle group. Use the cycle table and topology figure to decide whether to break the loop or record an explicit governance decision.

2Inspect spring-boot-build.core.spring-boot.main:org.springframework.boot

This package has the strongest combined structural signal in the report scope: Architectural cycle participant, fan-in 169 / fan-out 8.

Report Basis

SectionArchitecture Overview

Report basisWhole model

Active lensgraph

Structural intentNormal

FiltersWhole-model defaults; durable governance retained

Structural Map

Topology, cycle-only topology, and Matrix figures show broad packages structure and cycle evidence from complementary angles.

Rendered as Topology Open

How do packages and subsystems depend on each other in this report?

Configuration

Focus: project root
Filter: All visible packages
Heatmap: Off

Encoding

Nodes: Packages and package folders
Edges: Package dependencies; line width follows dependency weight
Border colour: Cycle, Safe Haven, intentional governance, and debt status from the model

Rendered as Dependency Matrix Open

Which packages depend on which other packages?

Configuration

Focus: project root
Filter: All visible packages
Heatmap: Off

Encoding

Rows: Origin packages
Columns: Target packages
Cells: Dependency count and cycle/heatmap styling where applicable

⚠️

Matrix hidden — 324 packages

The visible graph has 324 packages, which exceeds the safety and report-legibility limit of 100. Rendering a matrix this large may hang the browser or become unreadable when fitted to a report page. Raise the limit in Settings → Matrix → Max packages to render, or use the Focus filter to narrow the graph first.

Hierarchy & Size

Composition figures show how code size and packages are distributed through the selected scope.

Rendered as Composition · Sunburst Open

Where is the analysed codebase concentrated across the package hierarchy?

Configuration

Focus: project root
View: Sunburst
Size metric: Lines of code
Heatmap: Off

Encoding

Arc: Package hierarchy segment
Arc size: Relative lines of code
Colour: Hierarchy depth

Rendered as Composition · Circle Pack Open

How does the same hierarchy read as nested containment rather than radial depth?

Configuration

Focus: project root
View: Circle Pack
Size metric: Lines of code
Heatmap: Off

Encoding

Circle: Package hierarchy segment
Area: Relative lines of code
Colour: Hierarchy depth

Governance Summary

Repository documentReady

Revisionmissing

EvidenceFresh

Local view exclusionsNone

Safe Havens0

Intentional records0

Debt records0

Maintenance findings0

CoverageNot loaded

Metric Notes

CCD, ACD, RACD, and NCCD describe structural graph shape. Package tables include fan-in, fan-out, instability, and line count. Dependency scope: compiled JVM dependencies. ARV is the unweighted average of package Relative Visibility; GRV is public JVM classes divided by all JVM classes across the graph. They describe public-surface breadth, not actual reachability or a universal architecture-health target.