Skip to content

Studio — Execution & Artifacts

The execution engine and artifact manager. Where work gets done. Every AI-generated artifact is tracked, attributed, and cryptographically verified. Like the filesystem and application runtime of an OS, Studio manages what gets created, where it lives, and proves it hasn’t been tampered with.

interface StudioCommandResult {
success: boolean;
command?: string;
output?: string;
artifacts?: ArtifactEntry[];
manifest?: ArtifactManifest;
}
interface ArtifactEntry {
id: string;
type: 'output' | 'manifest' | 'template' | 'artifact';
format: string;
content: string;
hash: string; // SHA-256
provenance: ArtifactProvenance;
}
interface WorkflowDefinition {
name: string;
steps: string[];
}
import { BaselineStudioSystem } from '@baselineos/studio';
const studio = new BaselineStudioSystem();
// Execute a command
const result = await studio.executeCommand('generate', { format: 'json' }, {});
// { success: true, command: 'generate', manifest: { ... } }
// Register and render templates
studio.registerTemplate({
id: 'hello',
name: 'Hello',
format: 'text',
template: 'Hello {{user.name}}',
});
studio.renderTemplate('hello', { user: { name: 'Amani' } });
// "Hello Amani"

The layer sweep runs three checks against Studio:

CheckWhat it validates
Artifact trackingSHA-256 provenance hashing is active for all outputs
Provenance chainArtifacts are linked to a task context (task ID present)
Output formatOutput format is valid and negotiable (JSON, YAML, text, table, CSV, HTML)