Skip to Content

2026.3

MCP, Pipeline Editor, Log Center, Meta Editors and more!

There are tooooo many updates, read them on your own or ask your agent to get you a summary 😉. But be careful, we've prepared cool visualizations for each section!

New icon set

We've started creating a new icon set for the plugin from scratch. This release redraws or adds 41 icons across the tool windows, the main toolbar, the editor and B2C Explorer, and every other icon follows in the next releases.

Before
2026.3
B2C Entry PointsTool window · New icon
41 of 41 in the new style
Point at an icon to compare it, click to flip it, or switch the whole set

Native ISML

ISML is now a native language of the IDE, with its own parser, indexes and checks, and no language server. Templates keep the IDE's HTML support.

pdict has the members the controller passes to res.render. Completion lists them with their types, and the documentation beside the list says which line sets each one
  • Scripting: pdict is typed from every render call of your controllers, SFRA view data and @pdict lines, and a member that several calls set has the union of their types. A render call with a middleware the plugin can't follow, a spread or a model passed to a helper still types the members it names, and an open template follows unsaved edits of its controller. ${...} and <isscript> get the Script API.
  • Pipelines: a template that an Interaction node or SendMail renders gets the pdict keys the pipeline binds before it: Assign To_n, Eval Result, system pipelet outputs typed from the pipelet, the loop element-key and Script node @output parameters. Call nodes and the callers of a start node count too, and Cmd+Click (Ctrl+Click) on the member opens the node that binds it.
  • Variables across templates: <isset> and <isloop> variables reach the templates you include, decorate or call as custom tags, as B2C Commerce renders them. Completion names the template that declares each one, and Cmd+Click (Ctrl+Click), Find Usages and Rename follow them from one template to the other. Template variable is never read checks every template that sees the variable.
  • Overridden copies: a template copy that nothing includes or renders, such as an integration cartridge's customerCard.isml that your site cartridge overrides, sees the variables that the templates including its path set. A variable that several templates set has each of their types.
  • Completion: all 32 tags with their attributes and values, template paths, custom tags, and live and postfix templates.
  • Navigation: Cmd+Click (Ctrl+Click) on template paths, routes, slots, static files, resource keys, custom tags and template variables. Gutter icons and Code Vision show which controllers, scripts and pipelines render a template, which templates include or decorate it and which copy it overrides. One click opens the target, or a chooser when there are several. Find Usages of a script export lists the templates that require it.
  • Editing: Smart Enter, Surround With, folding, the Structure view, Rename of template variables, Extract Include, Extract Custom Tag and Introduce template variable. With semantic highlighting on, each template variable gets its own rainbow color, free and while the IDE indexes. <isinclude>, <isdecorate> and custom tags have their own colors, and Inherit values from on the color page gives them the tag color again. (, [ and { typed in ${...} close right before its }, so ${require( becomes ${require()}.
  • Inspections: syntax, attributes, unresolved templates, routes and resources, output encoding and pdict members, with quick fixes. A blue note marks each ${} in a <script> block, which B2C evaluates on the server, and a quick fix writes a browser placeholder as ${'$'}{name}. A bare # in template text is an error, fixed with ${'#'}.
  • Formatting: an Indentation mode and a Full mode, with EditorConfig keys. Reformat Code and Optimize Imports escape each bare #.
  • Fast in large projects: 3,010 templates in 40 cartridges index in about a third of a second, and pdict is ready in under a second on a project with 1,400 templates.

The file type, highlighting, editing aids, formatting and the checks that read only the template are free. Completion, typed pdict, navigation and the checks that read other files require a valid plugin license or trial. See ISML.

Template Graph

Show Template Graph on the floating toolbar of an ISML template opens a graph of everything around it: the controllers, scripts and pipelines that render it, the templates that use it, and every template it includes, decorates with or calls as a custom tag.

Show Template Graph on the floating toolbar opens the controllers that render the template and every template it includes, in a column per cartridge of the site's cartridge path. Hover or select a file to light its paths, read them in the details panel, or find a file, route or tag
  • Cartridge columns: each cartridge of the site's cartridge path gets a column, in path order, so you see which copy of a path renders.
  • Lines: renders, includes, decorates, custom tags, paths chosen at run time and remote includes each have their own color and dashes, labelled with their routes and tags.
  • Trails: hover or select a file to light every way between it and the template. The details panel lists those paths, then the files it's used by and uses.
  • Search: find a file, route or tag, and step through the matches.
  • Moving around: zoom, pan, fit the graph, go back to the template with Escape, and double-click a file to open it where it renders the template.

The Template Graph requires a valid plugin license or trial. See Template Graph.

Native Digital Script

.ds files and the server .js files of your cartridges are now Digital Script, a native language with the syntax, scope and checks Rhino applies on B2C Commerce. The file type of server .js files stays JavaScript, so ESLint, Prettier and the debugger work as before.

With the mode at 19.10, for..of, template literals and arrow functions are errors that name the mode that runs them. Optional chaining runs in no B2C Commerce mode, so it is an error whatever the mode
  • Rhino syntax: for each, XML literals and E4X members, default xml namespace and type annotations parse as code. Script compatibility mode marks the syntax your code version can't run, with a fix to the older form.
  • Import calls: importPackage, importClass and importScript bind their names for completion, Cmd+Click (Ctrl+Click) and Find Usages, with script path completion, gutter icons and Code Vision. Optimize Imports removes unused and repeated calls, on demand or on the fly.
  • Scopes and types: the server scope and its host objects, type annotations, for each loop types, E4X, and typed pipelet, hook and job step parameters.
  • let and const: Rhino doesn't scope them as ES6 does, so two inspections offer var instead, and a const declared in a block resolves and is typed for the whole function, as it runs on the server.
  • Pipelets: pdict is typed from the pipelet header, and Script nodes of the Pipeline Editor show its dictionary rows.
  • Editing: File Structure, folding, quick documentation, live templates, a New Digital Script File dialog, intentions and colors.

Parsing, highlighting, folding, live templates and the syntax checks are free. Import calls, quick documentation, the intentions and most inspections require a valid plugin license or trial. See Digital Script.

Pipeline Editor

Yep, now you can throw Eclipse into 🗑️ and do all the pipelines work inside your favorite IDE! The work on pipelines editor was started in 2024 but finished only Q1 2026.

Select a node to edit it in the Details panel, or hold Shift and drag a rectangle to select several, as on a desktop, and move them together. Pinch on a trackpad, or hold Cmd or Ctrl and scroll, to zoom at the pointer, something the Eclipse pipeline editor never had. Set a breakpoint and the debugger centres each node it stops on

There is still room for improvements.


I know that there are still many projects based on pipelines in 2026, so further featues development will be based on the feedback and demand.

  • Edit the diagram, switch to the XML source view and save back to the same pipeline XML.
  • Zoom the way you zoom a map: pinch on a trackpad, or hold Cmd (Ctrl) and scroll.
  • While you debug, the graph centers the node the session stopped on and frames it in red.
  • A node's script, template or pipeline opens the way B2C Commerce resolves it: with a site scope selected, the copy on the site's cartridge path opens at once. Otherwise a chooser lists every cartridge that has the file, with a filter.

The visual editor requires an IDE runtime with JCEF support and a valid plugin license or trial. Without JCEF, use the XML source editor. See Pipeline Editor.

Pipeline Groups

Click Groups in the Pipeline Editor toolbar, or choose Navigate → Pipeline Groups…, to search every pipeline of your linked cartridges in one popup.

Pipeline Groups opens on the group of the pipeline you are in. Option+Up and Option+Down walk the groups, the search finds pipelines, groups, start nodes and cartridges, and Enter opens the pipeline at the start node that matched
  • One search across pipeline names, start nodes, routes such as Cart-RemoveProduct, groups, cartridges and types.
  • Enter opens the pipeline on the matched start node, Shift+Enteropens the XML source and Cmd+Enter (Ctrl+Enter) selects the file in the Project view. Alt+Up and Alt+Downswitch groups.

See Pipeline Groups.

Getting Started

Getting Started is a new setup page in an editor tab. It takes a new or freshly cloned project from install to the first upload in one place, so you don't have to look for the license dialog, the credentials, the connection editor, cartridge linking and the site scope picker on your own.

Click Setup in the status bar to open Getting Started on the first open step. Each button runs the real action in the IDE, and the guide moves to the next step as soon as one is done. Once all five tracked steps are done, the item reads just Setup
  • Nine steps in one tab: activate the plugin or start the trial, add B2C credentials, add a connection, link cartridges, select a site scope, turn on upload, show the B2C project view, open the Log Center and connect AI agents to B2C MCP. Each button runs the plugin's own action, such as the license dialog, New Account, the connections editor, Find New Cartridges, B2C Site Scopes, the Log Center tool window or Connect Agent, and a short film shows what each step does.
  • Knows where you are: finished steps are checked off as you set things up anywhere in the IDE, and the page opens on the first unfinished one. The status bar shows Setup 2/5 until the five tracked steps are done, then Setup. Tick Show only the panda in the status bar to drop its text.
  • Sets up SFCC projects only: after indexing it opens by itself each time you open an SFCC project whose setup is unfinished, such as a freshly cloned storefront with cartridge folders not linked yet. A project without SFCC code gets no tab, no Setup item and no SFCC tool windows, only Find New Cartridges and More. Tick Don't open automatically to turn the tab off in every project.
  • A tour after setup: it then shows the main toolbar, the tool windows, B2C Explorer with Open in Log Center on its log files, the Project view menu and B2C Entry Points.
  • Offline: the page ships inside the plugin, so it works offline and matches the plugin version. Without JCEF, the same routes open the setup docs instead.

Open it again at any time:

RouteWhat to do
Status barClick Setup 2/5, or Setup once setup is done
Help menuHelp → Intellij SFCC Getting Started
Main toolbarMore → Getting Started
Find ActionCmd+Shift+A (Ctrl+Shift+A), type Getting Started

See Getting Started.

B2C Entry Points

The new B2C Entry Points tool window lists everything a request, a job or a hook can enter in your linked cartridges: pipelines, controller routes, hooks, step types and custom API endpoints. One search field covers all five.

B2C Entry Points has a tab for each kind of entry point in your linked cartridges, with its count. Each tab's tools sit in the search row: Controllers lists endpoints, each with its definitions in cartridge path order, or A to Z sections whose header stays pinned while you scroll. One search recounts every tab as you type, Filter Hooks picks a hook type or area, the context menu opens a registration, and a double-click opens the code
  • A tab for each kind with its count and a tree grouped by cartridge, pipeline group, controller or API.
  • Controllers show as Endpoints, each opening to its definitions in cartridge path order, as A to Z sections whose headers stay pinned while you scroll, or by cartridge. Hooks have Filter Hooks for type and area and group by cartridge or script. Each tab's tools sit in the search row, the one in use pressed.
  • Double-clicking a row opens the code that runs. Context menus open the hooks.json, steptypes.json, api.json and schema.yaml declarations.
  • Routes are found as the server registers them: server.get, post, use, http, https and include, aliases of require('server') and SiteGenesis guard.ensure guards. Hooks include the hooks file that a cartridge's package.json names.
  • Search Everywhere gets a B2C Entry Points tab that replaces the SFCC Routes and SFCC Hooks groups, see Search Everywhere.

See B2C Entry Points.

Code Vision

Code Vision lenses link routes, hooks and step types to their declarations, usages and endpoints. They come from native indexes, so they need no Node.js and no external process.

Above a route, lenses open or copy its storefront URL, list its usages and the cartridges that extend it. Hook and step type lenses open their declarations, custom API gutter icons and request hints link api.json, the schema and the script, and the lens in site.xml edits the cartridge path
  • Route lenses above every controller route and pipeline start node: Open Endpoint, Copy Endpoint URL, the usage count, and the SFRA prepend, append and replace counts.
  • Hook functions show their hook and step type functions their steptypes.json key. Gutter icons link the api.json , schema.yaml and script of a custom API to each other.
  • Jobs and Page Designer: a step type in steptypes.json shows used by N jobs, and a Page Designer type's JSON used by N pages or used by N components. A click opens the job step or the library content in its editor, or a popup when there are several.
  • Each lens group can be turned off in Settings → Editor → Inlay Hints → Code vision, including the new SFCC Page Designer group.

See Code Vision.

SFCC Route Calls

If your project wraps URLUtils in a URL helper of its own, add the helper on Settings → Languages & Frameworks → SFCC Route Calls: its string argument then completes routes, Cmd+Click (Ctrl+Click) opens the route, and each call counts in the usages of the route lens.

Add your own URL helper on Settings › Languages & Frameworks › SFCC Route Calls. Its calls then complete routes, Cmd+Click to them and count in the usages of the route lenses, next to the built-in URLUtils and URLAction calls
  • A rule names an optional class, one or more methods and the zero-based index of the route argument, for example getUrl with argument 0 for urlHelper.getUrl('Account-Show').Rules work in scripts and in ISML ${...} expressions.
  • Rules from intellij-sfcc.completions.routes in the project's .vscode/settings.json apply too.

See SFCC Route Calls.

Real-time script navigation

Completion and Cmd+Click (Ctrl+Click) in scripts follow the cartridge path the way the server does, through every module.superModule override.

A require that lands on an override still knows the whole chain. Completion lists the members of every copy down the cartridge path, and Cmd+Click follows each module.superModule to the copy that defines the member, or to the override that adds it
  • Members through overrides: a require that lands on an SFRA override completes the members of every copy down the cartridge path, and Cmd+Click (Ctrl+Click) on a member opens the copy that defines it.
  • Always current: a script added, moved or deleted in another cartridge shows up in require completion and navigation right away.
  • The site's own copy: with a site selected, a */ path written with its .js extension opens the copy on that site's cartridge path.
  • Faster in large projects: hooks, site preferences, custom objects and the B2C Script API tables are kept until the files they come from change.

See Members through every override.

Resource bundle messages

Resource.msg and Resource.msgf calls now read as the text their bundle holds,and a module path kept in the require bundle works like a path written in the code.

A module path kept in the require bundle folds to the path the bundle holds. Cmd+Click on require opens that script, and completion lists its exports with the cartridge they come from
  • Messages as text: a call with a literal key and bundle folds to its message when the file opens, in scripts, Digital Script, TypeScript, ${...} and <isscript>. The default locale's file wins, then the first cartridge on the cartridge path. msgf shows its arguments in place, such as {pdict.points}.
  • Open and fold back: rest the pointer on the text to see the call, click the text to open it, and click the marker in the gutter to fold it again. Turn it off with Resource.msg() messages in Settings → Editor → General → Code Folding.
  • Module paths: require(Resource.msg(key, 'require', null)) resolves like a literal require . Cmd+Click (Ctrl+Click) on require opens the script, its exports are typed and complete with the cartridge they come from, and Cmd+Click (Ctrl+Click) on the key still opens its property. A key no bundle has falls back to the default, or to the key itself, as on the server, so a key that is a script path resolves and folds to that path.

Folding and module paths require a valid plugin license or trial. See Paths from a resource bundle.

Typed forms

Forms are now typed from their XML definitions, with no generated files: SFRA server.forms.getForm('profile'), dw.web session.forms.profile, session.getForms() and pdict.CurrentForms.profile, and SiteGenesis app.getForm('profile.customer') and FormModel.get(...). TypeScript cartridge files get the same features.

A form from server.forms.getForm is typed from its XML. Completion lists its fields with their type, label and cartridge, and Cmd+Click on a field opens it in the form, where each node shows its usages and each message key its text
  • The copy in use: types come from the default copy in the first cartridge on the site's cartridge path, includes expanded, and flow into ISML pdict. That copy wins the whole file: copies are never merged.
  • Completion and navigation: fields complete with their type, label and cartridge, ahead of the base members, and Cmd+Click (Ctrl+Click) opens them in the form XML. SiteGenesis handlers such as handleAction are typed, and FormModelOf<'profile.customer'> names a form in JSDoc.
  • Find Usages and Rename: from a formid or any member read, across scripts, templates and pipelines. dwfrm_ HTML names resolve, complete and follow a rename too, in script strings, jQuery selectors,ISML attributes and req.form .
  • Search and copy: Go to Symbol finds form names, dotted paths and dwfrm_ names, and Copy HTML Name copies a field's dwfrm_ name. Quick documentation shows a member's form node.
  • Form XML: usage counts in Code Vision, the text of each message key inline, gutter icons for overriding, overridden and locale copies with Compare with Default Copy, Compare with Overridden Copy and Compare with Copy in Use, and a typed Structure view. object-def, binding and attribute-def resolve to the system object type and its attributes.
  • Inspections: a Forms group reports unknown forms and elements, with Create form, Change to and Add field, include cycles, duplicate elements, locale copies that differ from default and pipeline bindings that don't match the form.

Typed forms require a valid plugin license or trial. Schema validation of form XML stays free. See Forms.

Typed custom attributes

Custom attributes are typed from the metadata XML in your project, as in VS Code: Site.current.preferences.custom, and custom or getCustom() of products, orders, customers and every other system object.

Custom attributes complete with their display name and the type B2C Commerce returns, and getCustomPreferenceValue returns the preference's type. Cmd+Click opens the attribute in the metadata XML, where Show Usages lists every script, template and attribute group that reads it
  • The type B2C Commerce returns: string, number, boolean, Date, MarkupText, MediaFile, arrays for sets, and EnumValue with the allowed values for enums. Completion shows each attribute's display name.
  • Preferences: Site.getCurrent().getCustomPreferenceValue('id') returns the preference's type, and SitePreferenceOf<'id'> names it in JSDoc.
  • Everywhere a value goes: values typed with JSDoc, such as @param {dw.catalog.Product} product,and ISML <isset> variables that hold them.
  • Navigation and usages: Cmd+Click (Ctrl+Click) opens the attribute-id in the metadata XML. Find Usages from the definition or any read lists every script and template that reads the attribute and the form fields bound to it.
  • Usages in the metadata XML: each attribute-definition shows how many places read it, and a click on the count or Cmd+Click (Ctrl+Click) on the attribute-id opens them. Attributes nothing reads are grey.
  • Always current: types follow edits, changes on disk and branch switches at once.

Typed custom attributes require a valid plugin license or trial. See Custom Attributes.

Typed view data

res.getViewData() and res.viewData in SFRA controllers now have the type of what the route has set at that line, with no JSDoc: in a server.append, var viewData = res.getViewData() knows the base route's product and its type.

  • The route its own controller builds: the steps of the call's cartridge and of the copies below it on the site's cartridge path, each server.prepend before the steps it extends and each server.append after them, through every server.extend(module.superModule). A cartridge further left adds nothing, not even its prepends, since it runs only on the sites that have it. A step sees the steps before it and its own lines above the call. A route:BeforeComplete listener sees all of them.
  • Every way a step adds data: res.setViewData, the models of res.render, res.json and res.page, res.xml, members put on the object res.getViewData() returns and the middleware of the route, such as csrfProtection.generateToken with csrf.
  • Real types: each member has the type of the value that sets it, so viewData.product. completes the product model. A member set by several steps has the union of their types.
  • Navigation and usages: Cmd+Click (Ctrl+Click) on a member opens the line that sets it, or lists every step that sets it. Find Usages from a model property lists its view data reads in every cartridge.
  • Big objects fold: the type hint after var viewData shows {...} and opens the members on a click.
  • Always current: a step added, changed or removed in the call's cartridge or one below it changes the type at once, read from the index that types ISML pdict, so no other controller is parsed.

A helper the plugin can't follow, or a model it can't read, keeps the members it knows and accepts any other name. Typed view data requires a valid plugin license or trial. See Typed view data.

Metadata Editor

Metadata import files open in the Metadata Editor, with the XML on the Text tab beside it. The editor shows the file as a tree of object types, attribute groups and attributes, and every change is one undoable edit of the XML.

The Metadata Editor opens site archive metadata as a tree of types, groups and attributes, with a Text tab that keeps the caret on the selected row. Search and filter the attributes, edit display names in place, drag attributes into groups, add one from a gallery of types with a live preview, edit its properties and translations, list where code reads it, and see the properties and the XML side by side, where problems such as a repeated definition are listed and marked. The toolbar and the columns fold as the editor narrows
  • Toolbar: search by ID or display name, All, Custom and System, the No usages and Not in a group pills, and a Filters popup that counts what each filter shows. The problems count opens the XML with its problems.
  • Tree: display names, types and flags in columns, with the usages of each attribute. Double-click to edit in place, F2 to rename, and drag attributes into and between groups.
  • New Attribute: a gallery of the 16 attribute types with a sample value each, the settings of the type and a live preview of the Business Manager field and the script that reads it.
  • Properties: general settings, flags, enum values and the type's validation settings, with translations per locale. Copy Script Access copies how code reads the attribute.
  • Usages: a click on an attribute's count lists the scripts and templates that read it, and Delete says what still reads the attribute before it removes it.
  • Groups and types: a group lists its members in Business Manager's order, and a custom object type shows its data scope, replication, retention and key.
  • XML and problems: the IDE's editor on the same file beside the tree, with repeated definitions, wrong defaults and schema errors listed and marked in the gutter. Delete This Definition removes a repeated one.
  • Split view: the properties and the XML side by side, the selected element marked in the XML. A click in the XML selects its row, and the Text tab opens on the selected row.
  • Valid XML: elements go where metadata.xsd puts them, and a flag turned off is written false, so an import turns it off on the instance too.
  • New › Metadata File creates system-objecttype-extensions.xml or custom-objecttype-definitions.xml.

Editing requires a valid plugin license or trial. See Metadata Editor.

B2C Metadata

The new B2C Metadata tool window on the right bar lists the site archive XML of your project: metadata, catalogs, price books, libraries, preferences, jobs, services, slots and the other import files.

B2C Metadata lists every site archive XML file of the project under Organization and Sites, with its size and archive. Narrow it to one site, list the kinds first with Group by Type, open a file into what it defines with its usages in code, keep only what no code names, search by name, kind or path and open a file with Enter, copy its path or namespace, and find the same tree under META in the B2C project view
  • Organization and sites: each file sits under the organization or the site it belongs to, with its size and its archive. Group by Type lists the kinds first.
  • Scope and search: All, Organization, All Sites or one site, and a search that matches file names, kinds and paths.
  • Open: metadata files open in the Metadata Editor and the other kinds in their site archive editor. Right-click a file for Open Schema, Reveal in Project View and the copy actions.
  • META: the B2C project view shows the same files under META, beside the site's root.

New › Site Archive File creates an empty services, jobs, preferences, custom objects, slots or library file that imports without errors, the kind picked from the folder.

Right-click a folder and choose New › Site Archive File. The kind the folder holds comes first, and the file opens in its editor with its namespace and an empty root, ready for the first row. A name left as a kind's name follows the kind you pick

Fast on big archives: the window reads each file only up to its root element and keeps what it read across restarts, so a catalog of hundreds of megabytes costs what a small file does. A file shows within a moment of an edit, unsaved text included, and cartridge, node_modules and dot folders are skipped.

B2C Metadata reads each file only up to its root element and keeps what it read, so after a restart the whole tree is back as soon as indexing ends, the 412 MB catalog included. A file shows within a moment of getting a root element, before it is saved, and XML in dot folders, cartridge folders and node_modules stays out

See B2C Metadata.

Site Archive Editors

Services, jobs, site preferences, custom objects, content slots and content libraries open in the same design view as the Metadata Editor: a toolbar, a tree of what the file defines, and a form or the XML for the selected row.

The site archive editors open services, jobs, site preferences, custom objects, content slots and content libraries in the design view of the Metadata Editor. Search, narrow by scope or a quick pill, open the Filters popup, jump from the problems count to the XML and fix it with Alt+Enter, see the properties and the XML side by side, hide columns and turn grouping off, then edit each kind in its form: copy a createService call, create a missing profile, add and schedule job steps, copy and move a preference across tiers, define a preference, paste and add custom objects, make a slot configuration the default, write a missing Page Designer type and find the code that uses a row
  • Services: services, profiles and credentials with their links. Not linked and Broken links pills, Add Service, Copy createService Call and Create Missing Profile.
  • Jobs: flows, splits and steps with their schedule. Step parameters come from the step type, standard steps included, with Bind to Job Parameter… and Add Missing Parameters.
  • Site preferences: a column per instance type, = for the same as Development, Copy Development to Staging and Production, Compare Tiers… and Create Definition….
  • Custom objects: a column per attribute of the type, edited in place, and rows pasted from a spreadsheet.
  • Content slots: On and Default ticked in the tree and the template picked from its cell, content assets checked against your libraries, a schedule set with the IDE's date picker and Weekdays links, Add Configuration step by step, campaign assignments, Duplicate… and F2 rename.
  • Content libraries: the HTML body in the IDE's own editor beside, below or in the form, with fold, Soft-Wrap, Reformat and Expand. Translations for every localized field, Compare Locales… and Copy to Locales…, content links that complete and resolve, Override for a Site…, and Page Designer pages with a data form per component and Create Page Designer Type for a type no cartridge defines.
  • Shared tools: schema and kind checks with fixes on Alt+Enter, Copy as XML, Compare With…, New Uninstall File, and a Code column that counts the scripts and templates that name each service, slot or content.

The Site Archive tab of Search Everywhere finds services, jobs, preferences, custom objects, slot configurations and content by ID, and Enter opens the file in its editor with that row selected.

The Site Archive tab of Search Everywhere finds what the site archive files define by ID, with the kind and the file of each row. Custom objects carry their type and slot configurations their slot and context. The funnel hides kinds, and Enter opens the file in its editor with the row selected

The editors require a valid plugin license or trial. See Site Archive Editors, the Library Editor, the Slots Editor and the Site Archive tab of Search Everywhere.

Content links in library and slot HTML, such as $url('Page-Show','cid','shipping-info')$ and images/hero.jpg?$staticlink$, now resolve.

In the HTML of a slot or a content asset, $ lists the link functions with their documentation, and the call lists the routes of your cartridges, then the keys the route reads and the content assets of your site archives. The route and the ID that resolve get the thin underline of a reference
  • Navigation: Cmd+Click (Ctrl+Click) opens the route, the content asset, the folder, the category, the product or the static file a link names.
  • Completion: routes with their cartridges, the keys of the route such as cid and pid, and the IDs of the project with where they come from.
  • Content link problems: a new inspection reports unclosed links, unknown routes and content assets, with fixes. HTML inspections stay quiet inside link values.
  • Route usages count the content links that call a route.

See Content links.

Site archive IDs in scripts

Service, credential, preference and custom object IDs in scripts link to the site archive XML that defines them.

Cmd+Click on the ID in createService or ServiceRegistry.get opens services.xml on the Services Editor tab with the service selected, and setCredentialID completes the credentials with the file each comes from. With the file on its Text tab, the jump lands on the ID in the XML. An ID that only Business Manager defines stays plain, with no error
  • LocalServiceRegistry.createService('acme.http.orders', ...) and ServiceRegistry.get('acme.http.orders') open the service in the Services Editor with its row selected, and setCredentialID the credential.
  • getCustomPreferenceValue('apiTimeout') and setCustomPreferenceValue open the preference, and CustomObjectMgr.getCustomObject('PaymentCountry', 'AT') the custom object.
  • Completion lists the IDs of the project with the file each one comes from. An ID the project doesn't define is no error.

See IDs in scripts.

Overrides and cartridge sites

The Override dialog is a grid of your cartridges, and every file of a cartridge off the site's cartridge path says which sites use it.

The Override dialog lists the linked cartridges in the order of the selected site's cartridge path, with the ones off it last. Hover a cell for its position and what Override does there, set the columns and the order from the gear, pick how the new file starts, then list every copy with Show Overrides
  • Override dialog: the cartridge list and check boxes are now a grid with a cell per cartridge, in cartridge path order, a filter, totals such as N on the X path and N have copies, and a legend.
  • Copies stand out: the cartridge that holds the file is green, bold and tagged Current. A cartridge that has a copy already is yellow and tagged Overrides when its copy comes before the file on the site's path, Overridden when it comes after, or Has a copy off the path. The tooltip says which of the two the site uses, and off the path a tag says how many sites use the cartridge.
  • Type to filter, Up and Down to move, and Enter or a double-click to override. The gear sets the columns, the order, and whether to show cartridges off the path, the legend and Start on the Last Picked Cartridge.
  • Start with picks an empty file, server.extend(module.superModule), module.superModule or the file's content. Override creates the file, or opens the copy when the cartridge has one.
  • Off-path banner on scripts: a script in a cartridge that the selected site doesn't use shows a banner, like templates do, with N sites use it, Switch site and Dismiss.
  • N sites use it opens the sites with that cartridge on their cartridge path. Picking one makes it the site scope, and Switch site opens B2C Site Scopes.
  • Show Overrides marks copies off the cartridge path, says which sites still use them, and adds Compare with for each copy.
  • Cartridge chip: a chip next to the inspection counter names the file's cartridge. Click it to select the cartridge folder.

See Overrides.

Site Scopes

B2C Site Scopes lists the sites of the active instance and of the local site.xml files. Hover a site for its cartridge path in columns, pick it to order code insight by that path, or open the path from the site's ⋮ button to filter it, copy it and reveal a cartridge
  • Hover card: resting on a site shows its name, ID, source and cartridge path at once.
  • Cartridge path popup: a filter, a Not linked tag on cartridges the project doesn't link, Copy path, and Enter or a double-click to reveal a cartridge. A legend counts the linked and not linked cartridges.
  • On staging and production instances, the popup says Site scopes are not available for X instances yet.

See Site Scopes.

Cartridge Path Editor

The Cartridge Path Editor opens the cartridge path of a local site.xml or Business Manager preferences.xml as an ordered list.

The lens on the cartridge path of a local site.xml opens the path as an ordered list. Drag entries into place or drag a linked cartridge in, and every change is an edit of the XML document that you save like any other file
  • Click the Edit cartridge path: RefArch (4) Code Vision lens above <custom-cartridges> or <preference preference-id="CustomCartridges">, or the cartridge icon in the gutter.
  • Drag entries to reorder them, drag in a cartridge linked to the project, type a name or choose Add All. Names that no linked cartridge has are marked Not linked to this project.
  • Each change is one undoable edit of the XML document, and only the cartridge path value changes.

The lens and the gutter icon require a valid plugin license or trial. See Cartridge Path Editor.

Compare with B2C Instance

Compare with B2C Instance shows how a file of a linked cartridge differs from its copy in the active code version, in the IDE's diff viewer, unsaved changes included.

Right-click a file in a linked cartridge and choose Compare with B2C Instance. The diff puts your local text, unsaved edits included, next to the copy in the active code version. Upload Local sends your side, and Take Remote replaces it as one edit that Undo reverts
  • Open it from the editor, editor tab or Project view context menu, the View menu, the editor floating toolbar, or with Cmd+Option+Shift+O (Ctrl+Alt+Shift+O).
  • Upload Local saves the file and uploads it to the active code version.
  • Take Remote replaces your local text with the copy on the instance as one undoable edit.

The compare requires a valid plugin license or trial. See Compare with the instance.

Switch to Active Code Version

A wand next to Select Code Version on the main toolbar asks the instance which code version is active and selects it, as the wand in the VS Code extension does. It takes the place of Refresh Code Versions.

The wand asks the instance which code version is active and selects it, so uploads go where the storefront runs
  • One click: the wand spins while the instance answers, then the selector shows the active version and a notification names it, or says it is selected already.
  • Uploads follow the storefront: uploads and Auto Upload go to the selected code version, so after someone activates another version in Business Manager, one click sends your changes there.
  • The list stays current: opening Select Code Version asks the instance for its latest versions and marks the active one [Active on instance].

See Switch to the active code version.

Instance colours

Every Production, Staging and Development connection wears its kind's colour, in Business Manager's hues: red, blue and green. POD sandboxes and On-Demand Sandboxes keep the plain cloud.

Production is red, Staging blue and Development green, in Business Manager's colours, and sandboxes stay plain. A group with any of them shows the PIG icon. The colour follows the instance into Select Connection, Deploy Code, B2C Sync Log, Log Center and the tabs of files opened from it
  • Read from the hostname: production-, staging-, cert.staging. and development- hosts get their kind. When the host doesn't say, a name such as PROD EU, stg or Dev 2 gives it.
  • Bands in B2C Explorer: each instance and the rows under it lie on a band of its light colour, so a log file or a code version always shows which instance it belongs to.
  • PIG groups: a group with a Production, Staging or Development connection shows the PIG icon on its own. Group Icon in the group's menu, or Icon on the group in Connections Settings, picks another: PIG, Group, Clouds, Folder, Classic or a kind's cloud.
  • Wherever the instance shows: the toolbar's Select Connection, Deploy Code targets, the instance's tabs in B2C Sync Log, B2C OCAPI and the Log Center, B2C Credentials, B2C MCP, and the editor tabs of files downloaded from the instance.

See Instance colours and PIG groups.

Deploy Code

Deploy Code is one dialog for one instance or many: check the instances, check the code version on each row and deploy to all of them at once, each into its own code version.

Click Deploy Code: the instances you checked last time come back checked. The wand on a row, or Switch All to Active Code Versions, picks the version the instance runs. A .apiversion file forces the code version to run on one B2C Commerce API version, for compatibility: a row's API version picks it for that instance, API version under the list for all, and Not set uploads none. Deploy opens B2C Sync on its Code Deployment tab, one tab per instance, and one notification sums up all of them with the time
  • Remembers your instances: the dialog checks the instances that were checked when it last closed, instead of only the active connection, and keeps folded sections folded for the project.
  • Sections and groups: the active connection comes first, then Other connections and each group with its icon and a box that checks all of it. Titles count the instances and the checked ones, such as · 2 selected, fold on a click, and show a problem inside a folded section. Expand All Groups and Collapse All Groups sit in the toolbar.
  • Filter instances: the field in the header narrows the list by name, hostname, code version or group, opens the sections that match and marks the matching text. Esc clears it before it closes the dialog.
  • A code version per row: each row starts on the version selected for its connection, as the main toolbar shows it, or on the one active on the instance. A status such as Active, Inactive or Missing says where the code lands. Click the version to pick another, switch to the active one or create a New Code Version..., with a search from 7 versions.
  • A wand per row: next to a row on another version, the wand switches it to the version active on the instance, and Switch All to Active Code Versions in the toolbar switches every row at once.
  • Notifications with times: "Deployed to 3 instances in 0:41" or "Deployment failed on 1 of 3 instances after 0:44", with one line per instance: the code version, the cartridges deployed, failed or skipped, the size and the time. Show Sync Log opens the instance that failed first.
  • Deployment Settings... opens from the dialog or the Tools menu. When an instance can't send its code versions, the list keeps the cached ones and a notification says so.

See Deploy Code.

B2C Sync Log

B2C Sync Log has two tabs beside the console, Code Deployment and Code Upload, and opens on the tabs alone.

B2C Sync opens on the active instance right away, with a tab on top for each instance in its colour. The buttons at the end of the tabs put Code Deployment and Code Upload side by side, each with its own toolbar, or bring the console in, and the ⋮ menu picks how uploads show in the editor
  • Pick the layout: Hide Console (the default), Show Console Beside and Show Console Only on its toolbar, kept for every project. Show in Console on a row brings the console in beside the tabs at that row's line.
  • A tab per instance: each connection a deployment or an upload went to has its own tab on top, with its own Code Deployment and Code Upload. The tab icons show a running badge, and a red one when something failed.
  • Links to B2C Explorer: the code version and the host in each summary open B2C Explorer with that code version or connection selected.

Code Deployment shows the last deployment as a row per cartridge, global modules folder and B2C module. A deployment that starts opens it, grows the tool window until the first rows show and puts the keyboard in the table.

Deploy Again runs the last deployment once more. The headline follows it from zipping to Deployed 12 cartridges in 0:41, the strip fills a segment per cartridge sized by its zip, and the running rows stay on top while they go up
  • One line on top: "Deploying to version2" with a spinner while it runs, then "Deployed 12 cartridges in 0:38", "11 deployed, 1 failed in 0:44" or "Stopped after 0:12: 5 of 12 deployed", followed by the version, host and size.
  • The strip: a segment per cartridge, as wide as its zip, that fills as it goes up and turns green or red. Hover a segment to name the cartridge, click it to select its row.
  • Every step of every cartridge: Queued, Removing old copy, Uploading 64%, Waiting for the instance, Unzipping, Cleaning up, then Deployed, Failed with the error, such as "521 Sandbox is stopped", or Skipped.
  • A table that stays put: running cartridges stay on top whichever way it sorts, it never scrolls by itself, and column widths are remembered. Pills with counts filter by state while it runs, and Filter cartridges (Cmd+F) by name.
  • Act on it: Deploy Again, Retry Failed and Skipped and Stop Deployment on the toolbar. A row's menu has Show in Console (Enter or a double-click), Select in Project View, Copy and Deploy Selected Again. A retry sends only those cartridges, without the tasks and the full cleanup.

Code Upload lists every file and folder that Auto Upload or Upload sent, removed, moved or renamed, newest first with running uploads on top.

Save a file with Auto Upload on and its row comes in on top: Uploading with the percentage sent, then Updated, how long it took and when. The headline counts what runs, the details link the code version and the host to B2C Explorer
  • What and how: Created, Updated, Removed, Moved from, Renamed from or the error in red, Auto or Manual, the size, how long it took and when. The summary says whether Auto Upload is on, or Uploading 3 files, with the version and host links, the totals and how long the history is kept.
  • Find any upload: pills with counts for Running, Uploaded, Removed, Moved and Failed, Filter files by path, and a Filter Uploads button for the latest upload of each file, uploads on save or by hand, and today's or the last 7 days'.
  • Row menu: Open in Editor, Compare with Instance, Upload Again, Show Uploads of This File, Show in Console, Select in Project View, Copy and Remove from History.
  • Toolbar: Upload Again, Retry Failed Uploads, Auto Upload, Keep History for 1, 7, 30 or 90 days, Options and Clear History.
  • Thousands of uploads: the history keeps up to 10,000 uploads per instance across IDE restarts. The table holds the newest 200 and takes 200 more each time you scroll to its end.

See B2C Sync Log, Code Upload and Follow a deployment.

Upload progress

A file that uploads shows it where you are looking: in its editor and on its row in the Project view.

Line at the Top of the Editor, the default: a thin line under the tab fills as the file goes up, turns green and fades once the instance has it, or turns red for a moment when the upload failed
  • In the editor: a thin line along the top of the editor fills as the file goes up, turns green and fades once the instance has it, or turns red for a moment when the upload failed. A folder's upload draws it in the editors of its files.
  • Your choice of style: Upload Progress in Editor, in the Options menu of Code Upload and the ⋮ menu of B2C Sync Log, picks the line (the default), a ring in the editor's floating toolbar that opens the upload in Code Upload on a click, or None, for every project.
The Project view follows every upload: a still spinner while a file waits its turn, a turning one while it goes up, with how far once it takes a while, and a green check that fades. A closed folder counts what goes up under it, and a failure stays as upload failed until the file goes up again
  • In the Project view: in the Project view, the Salesforce B2C Commerce view and the other project views, a still spinner marks a file that waits, a turning one a file that goes up, with how far for a large one, and a green check fades once the instance has it. A closed folder turns while files under it upload and says how many.
  • Failures stay visible: a failed upload shows a red mark at once and the row says upload failed, with the code version and the error in the tooltip, until the file goes up again. A folder says how many under it failed.
  • In the status bar: uploading a single file shows its progress for as long as it goes up, and Stop stops the request.

See Progress in the editor and Progress in the Project view.

Log Center

Log Center parses your instances' logs into a store on your computer, so you search, group and chart them in the IDE while you keep coding.

The Log Center button on the left stripe opens its home page, where Open in Tab opens an instance with Files, Issues, Events, Anomalies, Monitor and Storage, each with its count. The + lists the instances, or groups them by kind. The search filters the issues as you type, and suggestions list recent searches and the controllers that match. Picking a suggestion turns it into a filter chip, here Controller: Cart-AddProduct. The time range list narrows everything to a period, here Last 24 hours. Filters counts each level, file type and site under the current query, and a tick adds a chip. The Events chart marks code deployed from the IDE, and dragging over it narrows the range. Events opens in Unique, one row per kind of event with its count and when it happened, and As Recorded lists every entry
  • One tab per instance: Open Instance… opens any connection or stored instance in its own tab, coloured by kind. Tabs come back with the project, and closing one asks whether to keep its stored logs.
  • Six views: Files lists the instance's log files, Issues groups repeated errors and warnings, Events lists every entry under a chart, Anomalies lists what changed, Monitor follows what parses or is tailed and Storage shows where the logs are kept and their size. Each view's tab shows its count.
  • Your layout: move an instance to a window of its own or an editor tab, or maximize the Log Center. A short guide walks through the views the first time.
  • One query bar: search with *, ?, ^ and $, a time range from Last 15 minutes to Last 14 days or a custom one, and pills for Quick filters, Level, Request, HTTP, Blade, Session and Triage. Filters shows every filter with a count, from source and cartridge to request type and stack. Issues and Events share them.
  • Unique or As Recorded: Events folds each kind of event into one row with its count, or lists every entry.
  • Suggestions as you type: recent searches, then the sites, controllers and sessions that match, with their counts.
  • Event timeline: drag over the chart to narrow the range, zoom it, and click a level to show only its events. Marks show the deployments, activations and uploads this IDE sent and the instance's active code version, and open B2C Sync Log on a click.
Selecting an issue shows its details the full height beside the list: when it was seen, its actions, a summary and its tabs, here Stack. Hide Details leaves a Details strip that brings them back. Mark Fixed strikes an issue through, and Move to Backlog with Backlog hidden in the Triage pill takes one out of the list. Anomalies compares the last 24 hours with the 7 days before them and ranks what changed. A gutter icon marks a line that logged errors, and Open in Log Center selects its issue
  • One issue per pattern: messages that differ only in times, ids, UUIDs or numbers count as one issue.
  • Full-height details: the error text, where it was logged, its stack frames and facts, where it occurs, and tabs for the latest entry, its events, the log, the stack and the code it names. Open Source opens the script line, and every value has a card that filters or opens it.
  • Mark fixed and keep a backlog: fixed issues are struck through, and an issue that logs again after its fix shows up in Anomalies as a Regression.
  • Anomalies: the last hour, 24 hours, 7 days or your range compared with the period before. New errors, spikes, bursts, returned, spreading and stopped issues and instance-wide incidents, ranked Critical to Low, with the code change that came just before.
  • Errors in the editor: lines that logged errors or warnings get the level's icon in the gutter, with the count and when it last happened, and turn grey once you change the line. A click lists the line's issues and opens one in the Log Center, and Highlight lines with log errors lists them in the Problems tool window too.
  • Fix with an AI agent: with B2C MCP or the IDE's MCP Server plugin, agents search the Log Center, read issues and their events and mark them fixed. Investigate with MCP copies a prompt for one issue.
Parse reads the selected files once, archives too, and the row shows how far it got. A parsed file that grew offers Parse new, and Parse New Data reads only what every file gained. Tail parses a file, then follows its new lines, and the status bar shows what is tailed. The status bar widget opens Monitor, where the counts follow the new lines. Storage shows this instance's store a line per part: its size, events, files and time span, the space each file takes, how close it is to the limits and its folder with its links. Every instance's stored logs follow, with Delete All…
  • Parse or tail: Parse reads a log file or archive once, Tail reads it and then its new lines every second. Parse New Data reads only what grew since the last parse. Tails resume when the project opens again.
  • Pick files fast: check boxes, pills for type, prefix, custom log, blade, state, location and age, and views grouped by type or prefix.
  • Tail All: tail every log file of the instance at once.
  • In the status bar: the Log Center widget shows the stored events, a bar while files parse and a beating dot while they're tailed, coloured by the load. Its card lists every instance with its events, size and work, and a click opens the busy instance on Monitor.
  • Any log: Add Local File parses a log a colleague sent, a log open in the editor offers Open in Log Center, and B2C Explorer has Open in Log Center and Tail in Log Center on log files, grouped by type or prefix.
  • Tail in an editor tab: Tail in Editor Tab shows the new lines of a tailed file as they arrive.
  • Logs in the editor: SFCC logs open as SFCC Log files with level colours, folded stack traces, error marks, links from paths, routes and services to your code, quick documentation and a structure view.
  • Stays out of the way: parsing runs in the background and waits while the IDE indexes, in power save mode or when memory is low. The store keeps the logs within the age and size limits you set, and Manage Stored Logs… shows and deletes what each instance keeps.
Export saves the selected issue, the current filters or all events as HTML, JSON or Markdown. Create Report writes an HTML report of the current filters and opens it in the browser
  • Export: the selected issue or event, the current filters, the selected files or all events, as HTML, JSON or Markdown.
  • Reports: events, errors, investigation and anomalies reports, saved in the project and opened in your browser, with a chart, the largest patterns and the evidence.
  • Secrets masked: passwords, tokens, cookies and authorization headers are masked in exports, reports, copies and agent answers.

See Log Center, Search & Filters, Files & Tails, Issues & Anomalies, Export & Reports, Logs in the Editor and Status Bar & Gutter.

B2C MCP

B2C MCP is the plugin's own MCP server. It gives Claude Code, Codex, Cursor, GitHub Copilot, Junie and other AI agents what the IDE knows about your SFCC project: the cartridge path, types, routes, hooks, logs, the debugger and your instances. Agents start read only, and every change to an instance starts as a plan the server checks before it runs.

B2C MCP shows the server, the agents you connect, the tools they see and every request they make. Agents change an instance only with the tools you turn on
  • Connect Agent: sets up Claude Code, Codex, Cursor, GitHub Copilot, Junie, Gemini CLI, OpenCode, Windsurf or any other agent. Write its config file after a diff, run its own command right in the dialog, or copy a setup prompt for its chat. Config for puts the config in the project or, for All projects, in the agent's own settings, so the agent reaches this project from any folder. Each agent connects under its client ID with a client key of its own, so you can cut off one agent without the others, or renew every client key when a config was shared. How client keys work shows both in a short animation.
  • Proof from the IDE's indexes: the agent opens with intellijsfcc_project_context, then proves types, finds every copy of a file along the cartridge path and checks its edits against the same indexes the editor uses.
  • Ten toolsets and presets: pick SFCC read only, SFCC operational, SFCC with plans or SFCC full at the top of the Tools tab, and turn single tools on and off under it. Block Writes Now leaves agents read tools only, at once.
  • Plan, then run: uploads, deploys, code version changes and WebDAV writes are plans. intellijsfcc_mutation_execute runs a plan only while the agent's tools and instances allow it, and the audit log keeps every write. Every upload, deploy, delete and activation shows in B2C Sync Log.
  • Tool access, global and per agent: the preset and tool checkboxes on the Tools tab apply to every agent, and Connect Agent can narrow one agent's tools and instances. Production takes no code uploads, as code reaches it through replication from staging.
  • Check for errors: "Check the sandbox for errors" ranks the Log Center's open issues and finds the line behind each. Agents also parse and tail logs, compare before and after a deploy, find anomalies and mark issues fixed.
  • Live debugging: agents drive your script and pipeline debug session on the active connection, or capture a variable's value at a line in one call. A script breakpoint can take a condition or a log message.
  • Sandboxes, jobs and site archives: agents plan starting, stopping and creating On-Demand Sandboxes, running jobs, importing or exporting site archives and clearing the page cache with your cache job.
  • 23 SFCC skills: short guides that teach an agent the right tools in the right order. B2C Add Skill in the Project view and editor, Connect Agent, the Server tab and the Tools menu add them to the project in the agent's own format, after showing every file. An agent's first connection offers them too.
  • Tips and notifications: a short guide walks through the tool window the first time, and Show Guide in the gear menu runs it again. Notifications say when the server could not start, when an agent connects for the first time and when a destructive call was refused.
  • Activity and Analytics: every request with its agent, status, time and context cost, kept on disk and searchable, the agents connected now, and the tools agents use most.
  • In Getting Started: the new setup step Connect AI agents opens B2C MCP and Connect Agent.

Agents need a valid license or trial. See B2C MCP, Connect Agents, Tools & Presets, Plans & Safety and What Agents Can Do.

Find New Cartridges

New cartridge folders are announced in a notification of the project's own window instead of a dialog that blocked every open project.

When the project opens, a notification lists the cartridges and modules it found. Review… opens the dialog, where the vendor copy is ignored and Link links the rest
  • From the notification: Review… opens the list, Link All links everything when no name is taken, and Don't Ask Again turns the startup check off for the project.
  • Review: folders are grouped by parent with a checkbox per folder, one folder per name is checked, and notes mark names that are shared or already linked. Ignore and restore a row or a whole folder, filter, reveal and copy paths, and bring back ignored folders with Show ignored.
  • B2C modules too: the search suggests B2C modules and the global modules folder next to your cartridges, with their own icon and kind. Node tools, front-end projects and test fixtures are not suggested.
  • In the background: the search can be cancelled, and Link closes the window and confirms what it linked.
  • Cartridges and Modules: the dialog that lists linked roots, formerly Cartridges and B2C Modules Root Paths, shows each kind with its count on the left and its folders grouped by location, with a filter, a docs link per kind and a line that says what an add skipped or replaced.

See Find new cartridges and Review linked roots.

Faster in large projects

Indexes, caches and startup were reworked so the plugin keeps up with large storefronts while you type.

Each row runs Before and Now at the same speed on its own scale, so the blue bar stops where 2026.3 is done. Measured on a large storefront
Measured on a large storefrontBeforeNow
Index a 114,000-file storefront with every plugin index32 s24 s
Every pdict ready on a project with 1,400 templatesabout 6 sunder 1 s
Resource bundle keys after opening a project or switching branches13 s0.1 s
Highlighting after a character typed in the SFRA checkout template88 ms27 ms
pdict. completion after a render call changed20 ms7 ms
Highlighting a library that 100 scripts import, after one typed character118 ms85 ms
Reindex after saving a 410 KB system-objecttype-extensions.xml44 ms33 ms
Reindex after saving a controller35 ms31 ms
The hooks index of the 114,000-file storefront1.3 s0.45 s
Highlighting restarts for 20 ISML index refreshes in a row201
Find New Cartridges scan after startup, 6,000 folders11 ms1.5 ms
Opening a project without SFCC code44 ms0.2 ms
Save handling for a refresh of 400 changed files3.6 ms0.5 ms
  • Caches follow only the files they read: typing in one file no longer recomputes pdict, script imports and form lookups across the project, and typing that only moves code recomputes nothing.
  • Typing in templates: names such as URLUtils or pdict.order.totals.subTotal keep what they resolve to while you type. Only an edit that changes what they read, such as an <isset>,an <isscript> import or the @pdict comment, resolves them again.
  • Indexing in one pass: script indexes share one lexer pass per file, and metadata XML is read without building the XML tree. A library or exclude change rereads only the folders it moved instead of rebuilding every index.
  • Quiet in other projects: a project without SFCC code starts no plugin services until you link a cartridge or add a connection.
  • Uploads on save leave the IDE's write action and go out in order from one queue. A file saved twice before its upload started is uploaded once.

Other changes

A new release shows its What’s new card once and the update check offers to install it. More › All Release Notes opens the release history at any time
  • IDE compatibility: support for JetBrains IDEs 2026.3 EAP builds.
  • Error reporting: plugin exceptions and IDE freezes can be reported through JetBrains Marketplace. The Sentry integration, its startup consent popup and its settings are removed.
  • Release announcements: a native popup shows what is new, and More → Plugin → All Release Notes lists every release, see Updates & Notifications. Notices from the Intellij SFCC team stay in the IDE's Notifications history. Web pages opened in an IDE tab show their address and an Open in Browser action.
  • Faster startup: connection files, cartridge settings and credentials load in the background.
  • Route navigation: Cmd+Click (Ctrl+Click) on a route opens its controller again when the Site Scope is Global.
  • Go to Declaration: Cmd+Click (Ctrl+Click) opens the target of a hooks.json hook name, a steptypes.json function key, an importScript('cartridge:path') call, and the start-name-ref, ScriptFile and template values in pipeline XML.
  • Where a name comes from: completion items declared in another file of a linked cartridge show that cartridge on the right with its icon, and the path in the cartridge after the type, such as (scripts/util/urls.js). Goto popups, gutter targets and Search Everywhere show the cartridge too. Script API names show their package instead, such as dw.util, TopLevel, OCAPI or SFRA.
  • Clickable strings: template, require and importScript paths, static files, hooks.json and steptypes.json values, custom attribute names, content asset IDs and dwfrm_ names get a thin grey underline while they resolve. Change or hide it with Clickable path, route or key (underline) on the ISML and Digital Script color pages.
  • Hooks: hooks.json completion adds 23 dw.extensions.* hook points, 334 in all. The gutter icon of a platform hook opens its script at the function, and the function's parameters are typed, such as basket as dw.order.Basket. A function several hooks share takes all their types, exports. completes the hook functions a script doesn't export yet, and a return value the hook doesn't expect is reported, see Hook function parameters.
  • Custom APIs: api.json gets a bundled schema, and its endpoint, schema and implementation values link to the script and schema, with completion and red unresolved paths. A hint after each endpoint shows its requests, such as POST /loyalty, with the full URL on hover, and opens the operation on click, see Custom APIs.
  • Step types: function and the chunk keys such as read-function in steptypes.json get a gutter icon that opens the function in the step's module. Find Usages, the usage count and Rename of a step function reach steptypes.json, so renaming exports.run rewrites the step, and a step typed into an open steptypes.json counts at once, see Step types.
  • Page Designer: the JSON schemas ship with the plugin, so validation and completion work offline. Navigation and group completion work on Windows.
  • Trial and activation: a JetBrains Marketplace trial unlocks the plugin. If your Intellij SFCC account is deleted, plugin access turns off and a notification says so.
  • Auto Upload: changes made outside the IDE, such as a Git branch switch, are picked up by the IDE's own file watcher instead of a native library that worked only on some operating systems and CPUs.
  • Docs links: every help, How it works and documentation link in the plugin opens the matching page of these docs.
  • Tool window icons: new icons for B2C Explorer, B2C Credentials, B2C Entry Points, B2C Sync Log, B2C ODS and B2C OCAPI.
  • Main toolbar icons: the toolbar buttons match the tool window icons. The active connection is a green cloud and the others plain clouds, Deploy Code is a parcel, and Auto Upload shows a large check or cross, so its state reads at a glance.
  • Credentials: password fields stay disabled until the stored secret has loaded, so a slow keychain can't let you save an empty value over it.
  • Metadata: changing only the type of a system object attribute, or only a description, updates completion and types at once.
  • Modules folder: the Modules icon and Mark as Modules no longer appear on a cartridge's own scripts/modules folder. Only the modules folder beside your cartridges is the global one.
  • Colors: let is colored as a keyword in server scripts, Digital Script and ISML <isscript>, like var and const.
  • Schemas: 57 import and export schemas follow B2C Commerce 26.10, with the new channels, commerce feature state, event routing and storefronts areas. The standard job steps of 26.10 type the step parameters in jobs XML.
  • Templates: res.render(cond ? 'a' : 'b') types both templates, and a locale copy such as templates/ja_JP/... is typed from what renders its path.
  • superModule: matched by its exact path, case for case, on every operating system.
  • Tool windows: the right bar holds B2C Explorer, B2C Credentials, B2C Entry Points and B2C Metadata, and the bottom bar B2C Sync Log and B2C ODS, with B2C OCAPI beside them. Once per project the update moves the plugin's tool windows back into one block, a window you moved to another bar stays there, and Window → Layouts → Default keeps their buttons.
  • ISML navigation lists: the gutter and Code Vision lists of decorated, included, rendering and overriding templates, custom tag usages and importScript importers show the template path below its locale folder, the tag and its line such as <isdecorate> · line 3, and the cartridge, with speed search and Open in Find Tool Window.
  • API packages in completion: after dw. each package shows the package icon and what it holds in grey, such as 45 classes or 7 packages.
  • Connections: B2C Explorer groups show how many connections they hold.
  • Breakpoint options: the instance checks breakpoint conditions, and Evaluate and log prints its value in the debug console. With Suspend turned off, the request runs on.
  • B2C Explorer writes: deleting the active code version asks again, Delete on log files clears them to 0 bytes, and uploads, deletes and moves show in B2C Sync and wait while a deploy holds the instance.
  • Menus: Auto Upload in the Tools menu takes effect, Open IMPEX in Browser opens Impex, Cmd+Shift+X cleans the active code version on the macOS keymap, and names with _ show as written. SFCC files no longer show an empty Run entry, and the .intellij-sfcc/transpiled folder is never taken for a cartridge.
  • Fixes: B2C Explorer no longer hangs after a cancelled or failed upload, download, delete or move. Bulk File Upload sends each file once, skips only names already in the target folder and stops on Cancel. A breakpoint removed before the debugger connects no longer stops requests, and Evaluate keeps +, & and -. The Site Scope popup shows loading, empty and error states. The debugger no longer moves watches into the Variables view. Extract Template Script works in templates with <isloop> or <isset> variables.
Every tool window opens from its button. B2C Explorer, B2C Credentials, B2C Entry Points and B2C Metadata sit on the right bar, B2C Sync and B2C ODS low on the left bar, B2C OCAPI low on the right bar. ⇧⎋ hides the active one

This release also carries forward the 2026.2.1 fixes, such as debugger startup with linked cartridges without a selected Site Scope.

Updates

2026.3.7

Read the 2026.3.7 release notes.

2026.3.6

Read the 2026.3.6 release notes.

2026.3.5

Read the 2026.3.5 release notes.

2026.3.4

Read the 2026.3.4 release notes.

2026.3.3

Read the 2026.3.3 release notes.

2026.3.2

Read the 2026.3.2 release notes.

2026.3.1

Read the 2026.3.1 release notes.

Install and activate

Update Intellij SFCC through Settings → Plugins → Updates, then restart the IDE if prompted.

The gear next to Search Everywhere in the main toolbar opens Settings on the Plugins page. Search Marketplace, install Salesforce B2C Commerce (SFCC) and restart the IDE. The plugin's tool windows join the tool window bars, and B2C Explorer asks for access: Check Activation re-checks the license and opens the IDE's license dialog while there is none

2026.3 is a new paid major release. Your subscription or perpetual license must cover this release. A perpetual fallback license for an earlier major release does not unlock it.

On this page