[GH-ISSUE #42] Add managed-runtime, bytecode, and package ecosystem providers #31

Closed
opened 2026-10-15 18:20:14 +00:00 by bsakaguchi · 2 comments
Owner

Originally created by @morluto on GitHub (Jul 12, 2026).
Original GitHub issue: https://github.com/morluto/rea/issues/42

Goal

Extend REA beyond native machine code to common managed runtimes and packaged applications.

Initial families

  • JVM bytecode and JAR/WAR archives
  • .NET IL, assemblies, NuGet packages, and single-file applications
  • Python bytecode, wheels, zipapps, and frozen applications
  • Node.js packages, snapshots, bytecode, Electron bundles, and native addons
  • WebAssembly modules
  • Lua and other embedded scripting runtimes

Capabilities

  • identify_runtime
  • inventory_package
  • inspect_bytecode
  • decompile_managed_function
  • resolve_dependency_graph
  • trace_native_bridge
  • compare_packages
  • recover_embedded_resources

Requirements

  • Route semantic requests to runtime-specific providers.
  • Preserve raw bytecode or IL locations alongside normalized symbols.
  • Record package-lock and dependency provenance.
  • Detect generated, vendored, bundled, and application-owned code.
  • Treat decompiler output as derived evidence rather than original source.

Acceptance criteria

  • Representative fixtures for each enabled runtime have deterministic inventories.
  • Managed-to-native boundaries can be followed in both directions.
  • Package and version comparisons identify ownership and behavioral changes.
  • Missing runtime tooling is exposed through capability discovery, not runtime surprises.## Focused JavaScript and bundle work

#117 is the focused implementation issue for semantic analysis of JavaScript and packaged application bundles. It covers module/chunk recovery, bundler graphs, source-map validation, package provenance, Electron boundaries, native-addon/WebAssembly edges, deterministic Evidence v2 output, and bundle comparison.

This issue remains the umbrella for managed runtimes and package ecosystems. Work in #117 should establish reusable provider contracts where possible without prematurely generalizing JavaScript-specific parser or bundler behavior into the domain model.

Originally created by @morluto on GitHub (Jul 12, 2026). Original GitHub issue: https://github.com/morluto/rea/issues/42 ## Goal Extend REA beyond native machine code to common managed runtimes and packaged applications. ## Initial families - JVM bytecode and JAR/WAR archives - .NET IL, assemblies, NuGet packages, and single-file applications - Python bytecode, wheels, zipapps, and frozen applications - Node.js packages, snapshots, bytecode, Electron bundles, and native addons - WebAssembly modules - Lua and other embedded scripting runtimes ## Capabilities - identify_runtime - inventory_package - inspect_bytecode - decompile_managed_function - resolve_dependency_graph - trace_native_bridge - compare_packages - recover_embedded_resources ## Requirements - Route semantic requests to runtime-specific providers. - Preserve raw bytecode or IL locations alongside normalized symbols. - Record package-lock and dependency provenance. - Detect generated, vendored, bundled, and application-owned code. - Treat decompiler output as derived evidence rather than original source. ## Acceptance criteria - Representative fixtures for each enabled runtime have deterministic inventories. - Managed-to-native boundaries can be followed in both directions. - Package and version comparisons identify ownership and behavioral changes. - Missing runtime tooling is exposed through capability discovery, not runtime surprises.## Focused JavaScript and bundle work #117 is the focused implementation issue for semantic analysis of JavaScript and packaged application bundles. It covers module/chunk recovery, bundler graphs, source-map validation, package provenance, Electron boundaries, native-addon/WebAssembly edges, deterministic Evidence v2 output, and bundle comparison. This issue remains the umbrella for managed runtimes and package ecosystems. Work in #117 should establish reusable provider contracts where possible without prematurely generalizing JavaScript-specific parser or bundler behavior into the domain model.
Author
Owner

@morluto commented on GitHub (Jul 17, 2026):

The .NET execution-free provider and managed workflows are delivered, and JavaScript application/replay support is substantial. Remaining breadth is JVM/Python bytecode, WebAssembly, Lua, additional package ecosystems, and bidirectional native-bridge coverage. Update this umbrella around those residual families rather than describing managed support as absent.

<!-- gh-comment-id:5001702370 --> @morluto commented on GitHub (Jul 17, 2026): The .NET execution-free provider and managed workflows are delivered, and JavaScript application/replay support is substantial. Remaining breadth is JVM/Python bytecode, WebAssembly, Lua, additional package ecosystems, and bidirectional native-bridge coverage. Update this umbrella around those residual families rather than describing managed support as absent.
Author
Owner

@morluto commented on GitHub (Jul 20, 2026):

Split into focused child issues:

These replace the broad remaining scope in this issue. Closing this parent can happen once all children are closed.

<!-- gh-comment-id:5023444767 --> @morluto commented on GitHub (Jul 20, 2026): Split into focused child issues: - [ ] #367 These replace the broad remaining scope in this issue. Closing this parent can happen once all children are closed.
Sign in to join this conversation.
No milestone
No project
No assignees
1 participant
Notifications
Due date
The due date is invalid or out of range. Please use the format "yyyy-mm-dd".

No due date set.

Dependencies

No dependencies set.

Reference
morluto/rea#31
No description provided.