Skip to main content

Status and roadmap

PKGViz is currently released as 0.7.x alpha.

The project already has a coherent end-to-end workflow, but the word alpha is intentional: internal responsibilities are being modularized and the automation surface is still growing.

Available today​

Current repository behavior includes:

  • project and language detection
  • parser selection with retained language candidates/evidence
  • TypeScript, Java, C++, Python, Delphi, and Kotlin parsing
  • package dependency graph construction
  • vendor/intrinsic dependency semantics
  • dependency weights
  • cyclic-component detection
  • concrete import evidence for cycle edges
  • interactive Cytoscape visualization
  • package-depth and vendor filtering
  • multiple graph layouts
  • CLI JSON audit export
  • browser JSON/XML audit export

Architecture work in progress​

The repository is moving generic responsibilities into focused reusable packages instead of letting PKGViz become one monolithic analyzer.

That includes continued migration around:

  • graph algorithms
  • dependency/import analysis
  • language analyzers
  • reusable audit/rule contracts
  • CI/build-tool adapters

The rule for this work is behavioral equivalence first, extraction second.

Automation direction​

Audit rules are intended to be reusable outside the UI.

The first mandatory blocking rule defined by the project is cyclic-dependencies. Future integrations can then consume the same rule semantics from different environments—for example Node/TypeScript workflows or build-tool plugins—without duplicating the analyzer.

More languages and ecosystems​

The parser architecture already supports several ecosystems with different project layouts. Additional languages should be added as focused analyzers with representative fixtures and explicit project-root behavior rather than generic regex accumulation.

Open-source positioning​

PKGViz aims to make structural architecture analysis inspectable and automatable without requiring a proprietary architecture platform.

The strongest contribution areas are:

  • higher-fidelity dependency analysis
  • additional language support
  • richer architecture rules
  • CI/build integration
  • better evidence and diagnostics
  • large-graph UX and visualization
  • documentation and reproducible examples

Follow development on GitHub.