dashboard

name	dashboard
description	Build Metabase dashboards via the `mb` CLI — lay out dashcards on the 24-column grid (`{col,row,size_x,size_y}` math, per-chart default sizes) and turn cards into a filterable, cross-linked app. Covers wiring filters to cards (parameters + parameter_mappings), the field-filter vs. raw-variable target grammar, linked/cascading filters and their foreign-key requirement, cross-filtering, click-through, multi-series overlays, and tabs. The whole-array replace semantics live in `core`. Triggers — "build a dashboard from these cards", "my dashboard is squished into half the width", "wire a filter to these cards", "make a filter cascade", "click a bar to filter the other charts", "add a dashboard tab", "add a second series", "why isn't my filter showing".
allowed-tools	Read, Write, Edit, Bash, AskUserQuestion

Dashboard

A dashboard starts as cards on a grid; it becomes an app when filters drive the cards, charts cross-filter each other, and clicks navigate. This skill owns both: the grid layout and the interactive layer.

core owns the transport mechanics — the whole-array replace semantics (editing dashcards or parameters replaces the entire set; omitted entries are deleted; new cards use negative ids), update-dashcard for a single safe patch vs. update --body for a full replace, and the parameter-values verb. Read it first (mb skills get core). visualization owns each card's chart and the full click_behavior key catalog.

Inspect before you wire: mb dashboard get <id> --json hydrates parameters, dashcards, and tabs; mb dashboard cards <id> lists just the dashcards.

Layout: the grid is 24 columns — not 12

Every dashcard carries {col, row, size_x, size_y} in grid units: col is 0-indexed from the left edge, row grows downward, and col + size_x ≤ 24. Full-width is size_x: 24 — Metabase's per-chart default width of 12 is half a row. A layout authored on the usual 12-column web-grid assumption crams the whole dashboard into the left half of the viewport. The server stores whatever geometry you send — overlaps and gaps included, no auto-fix.

Default sizes (w×h): scalar/smartscalar 6×3, pie 12×8, table/pivot/object 12×9, waterfall 14×6, sankey 16×10, heading 24×1, text 12×3, every other chart 12×6.

The standard shape — a KPI row of scalars across the full 24, charts in halves or thirds below, wide tables full-width:

"dashcards": [
  { "id": -1, "card_id": 101, "col": 0,  "row": 0, "size_x": 6,  "size_y": 3 },  // 4 KPIs × 6 = 24
  { "id": -2, "card_id": 102, "col": 6,  "row": 0, "size_x": 6,  "size_y": 3 },
  { "id": -3,            
                
                  
               
                   

Sanity-check before sending: rows should fill to 24 and at least one card must end at col + size_x = 24. If nothing in the array crosses column 12, you've authored a 12-column layout — double every width.

The wiring loop: a filter is a parameter + a mapping per card

A dashboard filter is one entry in the dashboard's parameters array plus a parameter_mappings entry on every dashcard it should control. A parameter with no mapping is an inert widget — the most common "my filter does nothing" cause.

// dashboard.parameters — the widget
{ "id": "status", "name": "Status", "slug": "status", "type": "string/=" }

// on each target dashcard — bind the widget to a column of that card
{ "parameter_id": "status", "card_id": 42,
  "target": ["dimension", ["field", 1779, null]] }

id is a slug-like string you pick (e.g. order_status), unique within the dashboard — Metabase stores any non-blank string verbatim, so reuse the slug rather than guessing an opaque value. If you genuinely need an opaque id, mint one with mb uuid; never fabricate one.

type is a closed enum — an unlisted value is a hard parse error that echoes the allowed set back: string ops string/= string/!= string/contains string/does-not-contain string/starts-with string/ends-with; number ops number/= number/!= number/between number/>= number/<=; date date/single date/range date/relative date/month-year date/quarter-year date/all-options; location location/city location/state location/zip_code location/country; plus category, id, boolean/=, temporal-unit, and bare number/text/date/boolean.

The target grammar depends on what the card is:

Card's query	target
MBQL column	["dimension", ["field", <field-id>, null]]
Native field filter tag	["dimension", ["template-tag", "<tag>"]]
Native raw variable tag	["variable", ["template-tag", "<tag>"]]

(Field ids from table get <id> --include fields; native tags from native-sql.) Because editing replaces the whole set, adding a filter is read-modify-write: dashboard get <id> --json, append to parameters and to each dashcard's parameter_mappings, send the full arrays back. To wire one card without touching the rest, update-dashcard <dash-id> <dashcard-id> --body '{"parameter_mappings":[…]}'.

Choose the interaction

Four distinct mechanisms — pick by what the user wants clicking or filtering to do:

Want	Mechanism
One widget filters several cards	a dashboard parameter mapped to each card (above)
One filter's choices narrow another's	a linked filter (filteringParameters)
Clicking a chart filters the other charts	cross-filter click behavior
Clicking navigates to a question / dashboard / URL	link click behavior

Linked (cascading) filters

Make a child filter (City) show only values consistent with a parent (State) by listing the parent's id in the child parameter's filteringParameters:

{ "id": "city", "type": "category", "filteringParameters": ["state"] }

Two hard constraints, both from the same root: linked filters read table-metadata foreign keys only. They ignore joins defined inside a saved question or model. So the parent and child columns must be connected by a FK set in metadata — if the cascade shows values it shouldn't, the FK is missing (fix it via metadata, then retry). And filteringParameters is incompatible with a values_source_type of static-list or card — a custom value source overrides the cascade, so Metabase clears the link. Leave the child's value source live (omit it) for linked filtering to work.

Cross-filtering (click a chart to filter the rest)

Set the driver chart's whole-card click behavior to crossfilter, mapping the clicked value into a dashboard parameter; map that same parameter onto the follower cards normally. The driver stays unmapped to it (it emits the value; it doesn't consume it). click_behavior lives in the dashcard's visualization_settings (whole card) or column_settings[<col>].click_behavior (per column on tables), with camelCase keys:

{
  "click_behavior": {
    "type": "crossfilter",
    "parameterMapping": {
      "status": {
        "id": "status",
        "source": { "type": "column", "id": "STATUS", "name": "Status" },
        "target": { "type": "parameter", "id": "status" }
      }
    }
  }
}

Click-through (navigate)

{ "click_behavior": { "type": "link", "linkType": "dashboard",
    "targetId": 7, "parameterMapping": { … } } }

linkType is question / dashboard (carry a parameterMapping to pass the clicked context) or url (a linkTemplate like "https://app/orders/{{ORDER_ID}}", {{Column}} interpolated). The full click_behavior/parameterMapping key catalog is in visualization's settings reference — don't hand-author a complex one; build it once in the UI and copy it (mb dashboard get <id> --full --json). Note: a native-SQL card can't drill through — only cross-filter and link click behaviors work on it.

Series, tabs, value sources

Multi-series overlay — put several cards on one chart (line / area / bar only): the dashcard's series is an array of card ids in draw order ("series": [43, 51]). Sending it replaces the set; an empty array clears it. Patch with update-dashcard.

Tabs — tabs is [{ "name": "Overview", "position": 0 }, …]; a dashcard joins a tab via dashboard_tab_id. Creating tabs and cards together uses negative ids (per core): give a new tab id: -1, point new dashcards at dashboard_tab_id: -1, and the create/update response returns the real ids. A filter widget only appears on a tab if it's mapped to at least one card on that tab.

Filter value source (the dropdown behind a parameter) — omit values_source_type to pull live distinct values from the mapped column; "static-list" + values_source_config.values for a fixed list; "card" + {card_id, value_field, label_field} to source from a query. mb dashboard parameter-values <id> <param-id> [--query <substr>] previews what a widget will offer.

Gotchas

• Auto-connect / the missing map: a filter that "does nothing" or "won't show" is almost always unmapped, or mapped only to cards on another tab.
• Time-grouping parameters (temporal-unit) bind only to a datetime column in the query's last stage — add one after a time-bucketed summary and it can't attach.
• Required + default: a required: true parameter with no default blocks its cards until a value is chosen; give it a default for expensive queries you don't want running unfiltered.
• Whole-array replace: never send a partial parameters/dashcards array to update — you'll delete what you omit. Use update-dashcard for a single-card change.

Don't

• Don't lay out on a 12-column assumption — full-width is size_x: 24; a layout where no card crosses column 12 renders in the left half of the viewport.
• Don't declare a parameter and forget the per-card parameter_mappings — the widget won't filter anything.
• Don't expect a linked filter to work off a model/question join, a custom column, or with a static/card value source — it needs a metadata FK and a live source.
• Don't hand-author a complex click_behavior — copy a UI-built one.
• Don't add a second series to a pie/scalar/table — it's line/area/bar only.