Advanced Authoring & Tools

RuneSchema 0.8.0.0 provides authoring through Helpy's separate desktop workspace. It can help you find a loaded record, prepare an editable JSONC starter, check its references, and export it. Opening the page, searching, and building a preview do not install a mod or change the game data.

Open the tools

Helpy / RSMK 0.8.0.0

Open Helpy with F2. Item Service handles everyday browsing and delivery; Transmog & Creation, Recipes and Authoring Tools provide the authoring workspaces. Tools, Settings and Live Log have separate navigation entries. UI preferences remain in ui.jsonc; RuneSchema runtime settings use settings.ini.

The shared item bay defaults to 20 items per page in a 5 columns × 4 rows grid. Change the page size below the items (presets or a custom count from 1 to 100), then use Previous / Next. Page size changes preserve the first item's approximate position without clearing the cart. The catalog fetches only enough bounded windows to fill the requested page; it does not start a Full Scan. A partial catalog is marked “more available” rather than reporting a guessed total.

Search runs after a short typing pause or immediately on Enter. All categories filters item types. View contains Grid/List and optional column/row controls; the lower divider still supports free-height resizing. Narrow windows reduce columns and short windows scroll within the page. Preferences are shared across inventory and editor bays and saved in ui.jsonc (views.itemPageSize, itemGridColumns, itemGridRows, itemGridFitRows).

Choose an item, then Add to cart, or drag it into the compact cart. Give items is the only primary delivery action. Loadouts holds kit import/export; More holds list/player refresh and advanced scan tools. An empty cart does not reserve the height of a populated staging chest. Existing colors, scaling, list-view preferences and delivery confirmation semantics are retained.

Transmog & Creation separates source selection, customization and the resulting item into connected cards, with the item bay below. Inspect the selected source before creating a clone. Native property and JSON overrides are in a separate expandable section. The source item is not changed.

Recipes connects inputs, station/timing and outputs. Drag items from the bottom bay into either side, or select an item and use the add buttons. Quantities are editable; a recipe can have multiple outputs. Choose a native station row and, where applicable, a labeled category. Optional processing duration has minimum and maximum bounds in seconds. These cards stack vertically when the available width is limited.

Use Browse to search native recipe metadata, or paste an exact cooked RecipeData object path and choose Load. Saving an override changes only the authored ingredients, outputs and optional timing; it preserves the cooked recipe's identity. Adding a station is optional and does not remove its original station placements. New recipes require a compatible station. Save writes a RuneSchema /recipes definition and reports its file path; restart to apply it. It does not silently grant inventory items. Station artwork comes from native row metadata or a cooked preview-icon path, not an automatic web download.

Normal catalog queries combine loaded objects with mounted asset-registry metadata. Unloaded AI/resource candidates remain visible but need Inspect selected definition before spawning. Inspection loads only the selected class and checks its native ancestry; native NPC-copy restrictions and resource save-exclusion checks still apply. The coverage message identifies when registry metadata is unavailable. Unknown mod directories are not guessed to be resource or AI classes: authored /spawns entries and the explicit Full Scan remain additional discovery paths.

Drag an item into the cart, or right-click Add to cart, select the intended recipients and choose Give items. Selecting an item does not grant it. Cart editing pauses during delivery; entries leave the cart only after matching inventory-delta confirmations. Failed or unconfirmed entries stay staged for inspection, without automatic retries. Recipients can be removed individually or cleared without changing item quantities.

PowerLevel badges request the native game texture and display the item's actual native value, with a text fallback. Masterwork labels use native flags/tags; detailed tooltips retain the native MasterworkType identifier when supplied by the item. Item names and high PowerLevel do not imply masterwork quality.

Use Export kit to save the cart as JSON or commented JSONC. Import kit lists kits in the displayed runtime/Helpy/kits folder under RuneSchema's per-profile Saved state. Copy shared kits into that folder and choose Refresh kits. Selecting a file previews its name/count; Load into cart replaces the staged cart, not your inventory. Loaded-item metadata is refreshed on the game thread without mounting missing content. Unknown or unavailable items are checked again at delivery.

Exports use a versioned, data-only RuneSchemaHelpyKit format: a schema version, optional name and up to 64 unique cooked item paths with quantities 1..10000. Imports are limited to 128 KiB and do not contain scripts, automatic delivery, target-player selection or authority overrides. Existing exports require explicit replacement confirmation and retain a backup. The in-window kit library does not open a Windows file dialog.

{
  "SchemaVersion": 1,
  "Type": "RuneSchemaHelpyKit",
  "Name": "Starter sword",
  "Items": [
    {"Item": "/Game/Gameplay/Character/Player/Equipment/Held/Sword/ITEM_Sword_Bronze.ITEM_Sword_Bronze", "Count": 1}
  ]
}

Delivery uses the existing authoritative inventory grant and confirmation path. Auto-equip is disabled until an inventory-instance equip command and its native slot/result contract are mapped; importing a kit cannot enable it. Advanced mode's Inspect native equip / unlock contracts exports read-only function metadata without invoking those functions or editing saves.

RuneSchema authoring workflow

Open Authoring Tools from Helpy's navigation; the old engine-embedded settings page is no longer the authoring entry point. If authoring is disabled, review its runtime setting in Settings and follow the displayed restart or session-activation requirements. Engine-backed actions still need a ready runtime and their own safety checks.

Use the workflow selector for content/schema authoring, presets, traces, profiles, inspection and results. Offline save cleanup remains a separate, guarded workflow rather than an automatic step in creating a mod.

Create an editable starter

1. Choose Content & schemas in the workflow selector.
2. Choose a Content type, or use All Loaders for the complete loader list.
3. Choose the loader and Format. Basic starter is the default. Use $Patch to change supported existing content, $Clone to create a supported copy, or Fields reference for research only. Unsupported formats are disabled for that loader.
4. Enter search text and select a result. The page says whether it searches installed authored files or loaded game names and paths. An inactive mod's authored file can appear in search results; that does not activate the mod.
5. Select the fields you need and choose Build preview. For a new item clone, set the mod folder and new name and generate a permanent persistence ID once. Keep that ID unchanged after release.
6. Review the draft. Use Check changes and dependencies to check the proposed changes and loaded references. Resolve reported problems before using Export JSONC.
7. Put the exported file in the matching folder of your own mod, then restart and test it on a backed-up character.

Export editable reference JSONC and Fields reference are research outputs, not installable mods. The structural schema for the selected loader can be exported from Structural schema. Under All Loaders, Export structural loader schemas provides the full set.

Items, recipes, journal entries, quests, NPCs, spawns, and events have focused content choices. Player and nameplate pages include their own guided controls; the Items page also points to Helpy's Transmog & Creation workspace for item authoring.

Know what each stage is for

Stage	Use it to
Mod Authoring	Find records and export loader definitions.
Presets	Run supported diagnostic presets.
Traces	Narrow a live behavior or event path.
Inspector	Examine a target when a path or field is uncertain.
Results	Review completed tool output.
Save Cleanup	Perform a separate offline save operation only when you intend to.

Authoring tools cannot make an incorrectly cooked PAK valid, and a preview is not proof that a mod works in a world. Follow the first-mod guide and test first entry, return to menu, re-entry, and a full restart.