Skip to content

🛡️ elevata 2.0.0 - Architecture Control

Choose a tag to compare

View all tags
@ilona0925 ilona0925 released this 25 May 09:40
· 27 commits to main since this release
v2.0.0
98323ce

This major release introduces Architecture Control as a controlled, gated and auditable UI workflow for executing metadata-defined architecture.

Execution itself has already been part of elevata. What changes with v2.0.0 is that execution becomes available through a guided Architecture Control UI: scope-aware, approval-aware, previewed before execution, and recorded afterwards as deterministic audit evidence.

Architecture State
  ↓
Architecture Change Report
  ↓
Architecture Approval Artifact
  ↓
Execution Preview
  ↓
Controlled UI Execution
  ↓
Architecture Execution Record

✨ Highlights

🛡️ Architecture Control becomes the guided UI path for execution

Architecture Control is now a central, scope-aware workspace for reviewing, approving, previewing and triggering controlled load execution from the UI.

It supports:

  • all-dataset scopes
  • schema scopes
  • TargetDataset scopes
  • TargetDataset scopes with target-only execution

Users can inspect architecture reports, verify approval readiness, preview execution impact, run controlled UI-triggered loads and inspect the resulting execution record from one guided workflow.

🔎 Execution Preview before UI-triggered execution

Before a controlled UI execution can run, Architecture Control shows an Execution Preview.

The preview explains:

  • selected root datasets
  • full execution order
  • upstream dependencies
  • dependency mode
  • report fingerprint
  • approval reference
  • execution gate status
  • preview fingerprint

Large scopes remain readable through compact summaries, while complete selected roots and execution order can be expanded on demand.

✅ Approval-aware execution gates

Controlled UI execution is gated by architecture state and review status.

Execution is allowed when:

  • the scope has no architecture changes, or
  • the scope has a matching Approval Artifact, and
  • the report is not blocked by policy

Execution remains blocked when:

  • approval is required but missing
  • approval drift is detected
  • blocking policy decisions are present
  • Architecture Guard detects unsafe execution drift

Approval never overrides policy blockers or load-runner guardrails.

▶️ Controlled load execution from the UI

Architecture Control can now trigger constrained elevata_load --execute runs directly from the UI.

The UI execution path is constrained and scope-bound. It does not expose arbitrary CLI flags and does not replace the CLI. Instead, it provides a guided operational path for approved or no-change architecture scopes.

The controlled UI execution path keeps:

  • preflight validation
  • materialization policy checks
  • Architecture Guard enforcement
  • approval matching
  • dialect-owned SQL rendering

The UI also shows immediate running feedback during execution and captures command output for review.

🎯 Target-only execution for focused iteration

TargetDataset scopes now support target-only execution from Architecture Control.

This allows users to rerun only the selected TargetDataset when upstream data is already available — ideal for iterative work on Bizcore or Serving objects.

Target-only execution is explicit, visible and restricted to TargetDataset scopes.

The default UI execution path remains lineage-aware and includes dependencies.

📜 Captured output and diagnostics

Controlled UI execution results now include captured output and error diagnostics.

Architecture Control shows:

  • output tail
  • error tail
  • expandable full captured output
  • expandable full captured errors
  • command arguments
  • command options
  • execution status
  • execution message

Terminal color escape sequences are cleaned up for readable UI output.

🧾 Architecture Execution Records

Every controlled Architecture Control execution now produces a deterministic Architecture Execution Record.

Execution Records capture:

  • execution identifier
  • operator
  • started and finished timestamps
  • duration
  • status and message
  • Architecture Control scope
  • dependency mode
  • report fingerprint
  • approval identifier
  • preview fingerprint
  • command invocation metadata
  • output and error tails
  • output and error line counts
  • deterministic record fingerprint

Execution Records are stored as JSON artifacts under:

ELEVATA_ARCH_EXECUTION_DIR=.elevata/executions

This makes UI-triggered architecture execution auditable and traceable without turning Architecture Control into a general command center.

🔒 Governance and Determinism

v2.0.0 keeps the Architecture Runtime contract strict:

  • Architecture reports remain deterministic review artifacts
  • Approval Artifacts remain bound to exact report fingerprints
  • Execution Preview fingerprints bind scope, dependency mode and report state
  • UI-triggered execution remains protected by the load runner
  • Architecture Guard remains authoritative at execution time
  • Execution Records provide deterministic audit references
  • Target-only execution is explicit and limited to TargetDataset scopes
  • Full captured output remains inspectable in the UI while persisted execution records retain compact audit payloads

This preserves the core elevata principle:

Metadata defines architecture.
Architecture Control governs UI-triggered execution.
Execution Records preserve audit evidence.
SQL remains an artifact.

📘 Documentation

The documentation has been updated across the Architecture Runtime surface:

  • Architecture Control Plane
  • Architecture Overview
  • Load Execution Architecture
  • Schema Evolution
  • Determinism & Execution Semantics
  • Getting Started
  • Platform Strategy
  • README

The .env.example configuration now includes:

ELEVATA_ARCH_EXECUTION_DIR=.elevata/executions

🧪 Quality and Stability

This release adds and updates tests for:

  • Architecture Execution Preview
  • execution gate behavior
  • controlled UI execution command construction
  • target-only execution
  • controlled execution output capture
  • Architecture Execution Records
  • execution record JSON rendering and storage
  • Architecture Control view behavior
  • session-bound execution result handling
  • scope-aware menu configuration
  • shadow-compare Full Refresh suppression behavior

The UI flow was validated for TargetDataset and target-only controlled execution.

Why this matters

v2.0.0 turns Architecture Control into the guided operational UI of the Architecture Runtime.

Architecture is now:

defined → reviewed → approved → previewed → triggered from the UI → recorded

Execution was already part of elevata. v2.0.0 makes it controllable from Architecture Control, visible before it runs, and auditable after it completes.

Full Changelog: v1.9.0...v2.0.0

Assets 2
Loading