Scaffold Valtren AI extensions directly from VS Code, connect to a real Valtren organization, and browse live semantic tables and fields while you author extension logic.
This extension builds on the published scaffold CLI:
It is designed for teams who want a lightweight editor workflow without duplicating scaffold logic across the SDK, public examples, and Marketplace tooling.
This release adds a dedicated Valtren AI activity-bar workbench with:
Connection
- connect, disconnect, refresh semantics, and inspect the current org/version
Semantics
- browse semantic tables and fields from the connected Valtren org
- click a table or field to insert it into the active editor
- get autocomplete and hover docs while writing table and
table.field references
- open a lightweight schema explorer for a selected table
- insert ready-made snippets directly from the schema explorer
Uploaded Extensions
- browse uploaded org extensions
- open source previews
- approve, enable, disable, and smoke-test uploaded packages
You can still use the command palette, but the sidebar is now the fastest day-to-day workflow.
Commands
Valtren AI: Create Extension
- prompts for extension name and runtime
- opens a folder picker
- runs
npx create-valtren-extension ... in a VS Code terminal
Valtren AI: Connect to Organization
- stores your Valtren base URL and API token securely in VS Code SecretStorage
- validates the token against your organization
Valtren AI: Show Connected Organization
- shows the active org, roles, and quick actions
Valtren AI: Disconnect from Organization
- clears the saved token and cached semantic metadata
Valtren AI: Validate Current Extension
- detects the extension root, runtime, entry file, and packageability
Valtren AI: Package Current Extension
- creates an org-ready ZIP under
.valtren/dist/ in the current workspace
Valtren AI: Upload Extension ZIP to Valtren
- packages the current workspace and uploads it directly to the connected org
Valtren AI: List Uploaded Extensions
- shows uploaded org extension packages from the connected Valtren org
Valtren AI: Test Uploaded Extension
- runs the platform smoke test for an enabled uploaded extension
Valtren AI: Browse Uploaded Extension Source
- previews uploaded extension files from the connected Valtren org
Valtren AI: Approve Uploaded Extension
- approves a pending uploaded extension for activation
Valtren AI: Enable Uploaded Extension
- enables an approved uploaded extension in the connected org
Valtren AI: Disable Uploaded Extension
- disables an enabled uploaded extension in the connected org
Valtren AI: Browse Semantic Tables
- loads live semantic tables from your connected Valtren org
Valtren AI: Browse Semantic Fields
- loads live table.field references from your connected Valtren org
Valtren AI: Insert Semantic Table
- inserts the selected semantic table name into the active editor
Valtren AI: Insert Semantic Field
- inserts the selected
table.field reference into the active editor
Valtren AI: Refresh Semantic Cache
- refreshes semantic metadata from the platform
Valtren AI: Open Extension Examples
- opens the public examples repo in the browser
Semantic authoring assistance
Once you connect to a Valtren organization and refresh semantics, the extension now helps directly inside the editor:
- semantic table autocomplete
- semantic
table.field autocomplete after typing .
- hover docs for tables and fields
This makes it much easier to write extension logic against real business entities without constantly switching back to the browser.
You also get snippet-style suggestions such as:
table_nametable_name.field_name- quick field-reference snippets built from the first known field in a table
- semantic-aware pattern snippets for:
- risk rule
- entity mapper
- case summary
- timeline projection
The extension also adds inline diagnostics for unknown semantic references inside quoted strings, for example:
"sales_order_header""sales_order_header.net_value"
That helps catch typos while keeping noise low in regular code.
Secure connection model
- API tokens are stored in VS Code
SecretStorage, not in your workspace files.
- The extension currently uses:
POST /api/org/contextPOST /api/data-semantics/overviewPOST /api/platform/version
- Semantic data is cached locally in VS Code storage for faster authoring.
Why this extension stays lightweight
This extension does not re-implement scaffolding logic. It calls the published Valtren CLI so the SDK remains the source of truth, then layers secure platform connectivity and semantic browsing on top.
Current runtime templates
node-packorg-zip-nodeorg-zip-pythonsidecar-pythonsidecar-javasidecar-dotnet
Recommended flow
- Run
Valtren AI: Create Extension
- Run
Valtren AI: Connect to Organization
- Pick the runtime that matches your team and deployment model
- Use
Insert Semantic Table and Insert Semantic Field while writing logic
- Cross-check the public guides in valtren-ai/extension-examples
- Validate, smoke-test, and version the extension before production rollout
Current workbench scope
This release adds the first real Valtren Extension Workbench slice:
- secure org connection
- status-bar connection summary
- dedicated activity-bar workbench views for connection, semantics, and uploaded extensions
- local extension validation
- local extension ZIP packaging
- direct ZIP upload to a connected Valtren org
- uploaded extension listing
- uploaded extension smoke tests
- uploaded extension source browsing
- approve / enable / disable actions for uploaded extensions
- live semantic table browsing
- live semantic field browsing
- editor insertion commands
- semantic autocomplete and hover docs in the editor
- semantic snippets and schema explorer panels
- semantic-aware pattern snippets for common extension logic blocks
- inline diagnostics for unknown semantic references in quoted strings
Next steps will add:
- richer capability previews inside VS Code
Support
Versioning
- This extension follows semantic versioning.
- See CHANGELOG.md for release notes and upgrade guidance.
