Skip to main content

sedaro

The sedaro repository provides the programming framework that connects your content to the Sedaro Platform. Use it to define model types, read and write models, execute simulations, interact with platform services, and turn results into dashboards and reports.

See the Content Repositories overview for how the framework works with Kits and ModSim.

It contains two separately implemented libraries:

Source locationLibraryPurpose
sedaro_py/Python package sedaroModeling, simulation orchestration and results, platform clients, workflows, artifacts, visualization, and reporting.
sedaro/Rust crate sedaroA blocking HTTP client for Artifact and Workflow services, including authentication and token refresh.

The Python library is the main entry point for building simulation content and workflows. The Rust client supports integrations written in Rust; it does not expose the Python framework's full modeling and visualization API.

Build and exchange models​

sedaro.model provides the Python implementation of SedaroML:

  • Ontology, fields, and relationships describe model structure and configuration.
  • Validators and derivers check configuration and compute values during model preparation.
  • Initializers, state managers, and timestep definitions describe simulation behavior.
  • SML reads and writes .sml documents containing model instances, schemas, and authoring metadata for use in Blueprints.
  • Input-discovery and sampling support helps identify uncertain model inputs for studies.

Use these APIs when you need a new model type or custom modeling behavior. When an existing Kits component fits, compose or extend that component first. The framework supplies the modeling machinery; domain-specific choices such as a satellite, battery, or communications channel come from model libraries or your own code.

Run simulations and use their results​

sedaro.simulation provides Simulation and SimulationResult, along with study-planning and sampling types. These connect a prepared model to the simulator and make its output streams available for analysis and visualization. Kits adds higher-level Scenario and Study interfaces on top of these facilities.

The simulator and its build-time/runtime packages are dependencies of the framework. See the Simulator overview for the execution model and the repository's docs/gotchas/ guidance for implementation-specific constraints.

Build workflows and integrations​

sedaro.workflows supports typed Python workflow functions using the @workflow decorator, input/output dataclasses, and WorkflowContext. ArtifactInput and ArtifactOutput handle declared file inputs and outputs. A workflow can run locally, start a remote run, wait for completion, and retrieve typed outputs or compose child runs.

The Python runtime works together with a deployed workflow.yaml definition. The YAML declares the container environment, steps, dependencies, resources, and exposed inputs/outputs; Python implements the work each step performs. The repository also includes deployment abstractions and the sedaro-deploy command for managing declared workflow deployments.

For direct integration, sedaro.client includes synchronous and asynchronous Python service clients. Use these to resolve and transfer artifacts, manage workflow resources, and monitor executions. The Rust crate offers its own Artifact and Workflow service clients for Rust applications.

See Workflows for platform concepts and Getting Started for the definition-to-run lifecycle.

Produce dashboards, reports, and artifacts​

The framework supports several complementary outputs:

PackageUse
sedaro.artifact and sedaro.clientRead, write, and address versioned platform artifacts and series data.
sedaro.vlConstruct Sedaro Visualization Language (SVL) dashboards with charts, tables, statistics, and playback entities. Widgets can reference simulation series instead of embedding all samples in dashboard JSON.
sedaro.reportAssemble Markdown reports, plots, and rendered documents. PDF rendering requires Typst in the execution environment.
sedaro.diagramsGenerate diagram content for analysis and reporting.

A typical workflow reads input artifacts, prepares and simulates a model, computes analysis, and publishes output artifacts. An .svl dashboard describes the presentation; its referenced series artifacts hold the dynamic data. Keep both available when sharing or revisiting results. See Artifacts and Series Data Artifacts.