skenion-runtime 0.43.0

Skenion native runtime
Documentation

Skenion Runtime

Rust native runtime for graph compilation, scheduling, rendering, preview, output, plugin hosting, control, and telemetry.

Runtime internals live in a Cargo workspace until external consumers justify extraction.

Active Surface

The active runtime surface is a current 0.1 project loader, planner, session API, and local preview process manager. Unsupported graph/project/node contract versions are rejected with structured diagnostics.

It can validate and plan:

  • Skenion current 0.1 project JSON files
  • graph 0.1 documents resolved against node definition manifests
  • current 0.1 patch libraries and subpatch expansion
  • duplicate node and port ids
  • edge endpoint existence
  • output-to-input edge direction
  • type + rate compatibility
  • fan-in/fan-out policy checks
  • node kind/kindVersion resolution
  • graph port snapshots against authoritative node definitions
  • topological execution plan skeletons
  • cycle detection
  • deterministic dummy execution reports
  • a local session-driven preview process manager
cargo run -- validate-project --project path/to/project-0.1.json
cargo run -- plan --project path/to/project-0.1.json --format text
cargo run -- plan --project path/to/project-0.1.json --format json
cargo run -- run --project path/to/project-0.1.json --frames 2 --format json

The preview process manager is exposed through the Runtime session HTTP API. Standalone preview child commands consume prepared plan or preview-document artifacts and are not graph import or authoring APIs.

Status

Bootstrap repository for the Skenion project. Implementation follows the public architecture and release rules defined in EchoVisionLab/skenion.

License And Credit

This repository is licensed under the Apache License, Version 2.0.

Redistributions must preserve copyright, license, and NOTICE information as required by Apache-2.0. If Skenion helps your artwork, research, publication, installation, or tool, please credit Skenion and EchoVisionLab.