Examples

Every mapping on this page was run through the resolver the flow uses, against the data shown. The outputs are what the engine returned, including the awkward parts.

Building a message payload

Query, concatenation and a nested object in one mapping. $.…items.length counts an array; $.…items[*].sku collects a field from every element.

Flow data
Mapping
Result

{
  "trigger": { "event": "order.created", "order_id": "ord_88", "customer_id": "cus_12" },
  "GET_CUSTOMER": {
    "status": 200,
    "data": { "id": "cus_12", "first_name": "Dana", "last_name": "Reyes",
              "email": "dana@example.com", "tier": "premium" }
  },
  "GET_ORDER": {
    "status": 200,
    "data": {
      "id": "ord_88", "currency": "usd", "total": 184.97,
      "items": [
        { "sku": "A-1", "name": "Widget", "qty": 1, "price": 149.99 },
        { "sku": "B-2", "name": "Gadget", "qty": 2, "price": 17.49 }
      ]
    }
  }
}

{
  "to": "$.GET_CUSTOMER.data.email",
  "subject": "Your order |$.trigger.order_id| is confirmed",
  "body": {
    "greeting": "Hi |$.GET_CUSTOMER.data.first_name|,",
    "order_id": "$.GET_ORDER.data.id",
    "line_count": "$.GET_ORDER.data.items.length",
    "skus": "$.GET_ORDER.data.items[*].sku",
    "total": "$.GET_ORDER.data.total",
    "currency": "$.GET_ORDER.data.currency"
  },
  "channel": "email"
}

{
  "to": "dana@example.com",
  "subject": "Your order ord_88 is confirmed",
  "body": {
    "greeting": "Hi Dana,",
    "order_id": "ord_88",
    "line_count": 2,
    "skus": ["A-1", "B-2"],
    "total": 184.97,
    "currency": "usd"
  },
  "channel": "email"
}

total stays a number and line_count stays an integer. Only the two | strings become text.

Filtering a list

The same mapping, three filters, three different result shapes.

Flow data
Mapping
Result

{
  "GET_TICKETS": {
    "data": [
      { "id": "T-1", "priority": "high", "status": "open",   "assignee": "sam" },
      { "id": "T-2", "priority": "low",  "status": "open",   "assignee": "ada" },
      { "id": "T-3", "priority": "high", "status": "closed", "assignee": "sam" }
    ]
  }
}

{
  "open_ids":      "$.GET_TICKETS.data[?(@.status==='open')].id",
  "high_open_ids": "$.GET_TICKETS.data[?(@.status==='open' && @.priority==='high')].id",
  "urgent_ids":    "$.GET_TICKETS.data[?(@.priority==='urgent')].id",
  "count":         "$.GET_TICKETS.data.length"
}

{
  "open_ids": ["T-1", "T-2"],
  "high_open_ids": "T-1",
  "urgent_ids": [],
  "count": 3
}

high_open_ids is a bare string, not a one-item list, because exactly one row matched. Add a fourth high-priority open ticket and the same expression returns an array. A step reading this value has to handle a string, a list and [].

Note also what is not expressible here. [?(@.priority==='high' || @.priority==='urgent')] fails the step — the | characters are consumed as concatenation delimiters before JSONPath sees them. Two separate keys, or an Eval step, are the way round it.

Flattening a nested response

A JSONPath key is resolved like any other expression, so a record can be keyed by a value from the run.

Flow data
Mapping
Result

{
  "GET_PROFILE": {
    "data": {
      "account": { "id": "acc_9", "settings": { "notifications": { "email": true, "sms": false } } },
      "contact": { "email": "ops@example.com", "phone": null }
    }
  }
}

{
  "$.GET_PROFILE.data.contact.email": {
    "account_id": "$.GET_PROFILE.data.account.id",
    "email_notifications": "$.GET_PROFILE.data.account.settings.notifications.email",
    "sms_notifications": "$.GET_PROFILE.data.account.settings.notifications.sms",
    "phone": "$.GET_PROFILE.data.contact.phone|"
  }
}

{
  "ops@example.com": {
    "account_id": "acc_9",
    "email_notifications": true,
    "sms_notifications": false,
    "phone": "null"
  }
}

phone is the four-character string "null", not an empty string — the property exists and holds null, and a trailing | stringifies whatever it finds. Only a path that matches nothing produces "".

Searching a tree

.. finds a field at any depth. @root reaches back to the whole data object from inside a filter.

Flow data
Mapping
Result

{
  "trigger": { "region": "eu" },
  "GET_SITES": {
    "data": {
      "regions": [
        { "code": "eu", "sites": [
            { "name": "Leeds", "contact": { "email": "leeds@example.com" } },
            { "name": "Lyon",  "contact": { "email": "lyon@example.com" } } ] },
        { "code": "us", "sites": [
            { "name": "Austin", "contact": { "email": "austin@example.com" } } ] }
      ]
    }
  }
}

{
  "all_emails":   "$.GET_SITES..email",
  "region_sites": "$.GET_SITES.data.regions[?(@.code===@root.trigger.region)].sites[*].name",
  "first_site":   "$.GET_SITES.data.regions[0].sites[0].name"
}

{
  "all_emails": ["leeds@example.com", "lyon@example.com", "austin@example.com"],
  "region_sites": ["Leeds", "Lyon"],
  "first_site": "Leeds"
}

Writing $.trigger.region inside the filter instead of @root.trigger.region fails the step with $ is not defined.

Building a URL

Concatenation joins literal text with resolved values. $.static holds the flow’s Static Data and $.env holds the run’s own metadata.

Flow data
Mapping
Result

{
  "trigger": { "user_id": "USR-789", "q": "widget" },
  "static": { "api_base": "https://api.example.com" },
  "env": { "run_id": "run_4f2a", "auth_token": "…", "headers": {} }
}

{
  "url":    "$.static.api_base|/users/|$.trigger.user_id|/orders",
  "search": "$.static.api_base|/search?q=|$.trigger.q|&limit=25",
  "headers": { "X-Run-Id": "$.env.run_id" }
}

{
  "url": "https://api.example.com/users/USR-789/orders",
  "search": "https://api.example.com/search?q=widget&limit=25",
  "headers": { "X-Run-Id": "run_4f2a" }
}

What a missing value looks like

Flow data
Mapping
Result

{ "GET_USER": { "data": { "name": "Alice" } } }

{
  "name": "$.GET_USER.data.name",
  "phone": "$.GET_USER.data.phone",
  "phone_text": "$.GET_USER.data.phone|"
}

{
  "name": "Alice",
  "phone": [],
  "phone_text": ""
}

The key is never dropped. Without a trailing | a missing value is an empty array; with one it is an empty string.

Next steps

JSONPath

Every selector and filter that works

Concatenation

Building strings with |

Map Step

The step whose entire job is a mapping

Eval Step

Logic that JSONPath cannot express

Get Started

Spaces

Meetings

Forms

Assistants

Collaboration

Flows

Account Management

Advanced

Developers

Examples

Examples

Building a message payload

Filtering a list

Flattening a nested response

Searching a tree

Building a URL

What a missing value looks like

Next steps

JSONPath

Concatenation

Map Step

Eval Step

​Examples

​Building a message payload

​Filtering a list

​Flattening a nested response

​Searching a tree

​Building a URL

​What a missing value looks like

​Next steps

JSONPath

Concatenation

Map Step

Eval Step

Examples

Building a message payload

Filtering a list

Flattening a nested response

Searching a tree

Building a URL

What a missing value looks like

Next steps