How can I correctly pass currentNodeInputIndex and map additional fields using item linking in n8n?

Below is an explanation addressing your two points:

---

1. How to Pass currentNodeInputIndex

The method

  $("").itemMatching(currentNodeInputIndex)

expects a number as its parameter. That number represents the position (index) of the current input item that you're processing in your Code node. In practice, you pass the current loop index from your code as the argument.

For example, in a JavaScript Code node you can do:

JSCopied
const items = $input.all();

for (let i = 0; i < items.length; i++) {
  // Get the current input item.
  const currentItem = items[i].json;
  
  // Retrieve the corresponding linked item from "<node-name>".
  const linkedItem = $('Merge All Fields Code').itemMatching(i).json;
  
  // For instance, add the email from the linked item into the current item.
  currentItem.restoreEmail = linkedItem.email;
}

return items;

Here, the loop variable i is passed as currentNodeInputIndex to match the current item with its counterpart from the previous node.

---

2. Mapping Additional Fields from Linked Items

It sounds like your workflow involves several nodes:

• Merge All Fields Code: Uses item linking to combine/merge fields.
• Handle Both Node Data Code: Updates fields from “Merge All Fields Code” if returning items exist.
• LLM Chain: Returns only one field.
• Ai Resume Fields Node: Needs to map more fields, but they aren’t available.

The key is to ensure that every node returns the necessary linked metadata so subsequent nodes can use item linking. Some suggestions:

• Ensure Full Data Propagation:
  In your “Merge All Fields Code” node, make sure you return all the fields you’ll later need. When you update items in “Handle Both Node Data Code,” merge the extra fields from the linked input using the item linking method as described above. If your LLM Chain is only returning a single field, it might be overwriting or not including the additional fields from the earlier nodes.

• Merge Data Carefully:
  In your “Handle Both Node Data Code” node, consider modifying the code so that it:

  • Checks if the returning item (from LLM Chain) has additional data;
  • If not, it can fetch any missing fields from the linked item (using itemMatching on “Merge All Fields Code”).

  That way, the output passed on to “Ai Resume Fields” contains all the fields you need.

• Use Alternative Methods if Necessary:
  If there is ambiguity (such as multiple matching items), you might want to replace .item with methods like .first(), .last(), or .all()[index] to explicitly target the correct field.

---

Summary

• Passing currentNodeInputIndex:
  It is simply the numeric index (typically available from your loop variable) indicating which item is being processed.

• Mapping Additional Fields:
  You must adjust your nodes to ensure that all required fields are retained and merged during the workflow. This might mean modifying the “Handle Both Node Data Code” node to combine data from both the LLM Chain and the linked item from “Merge All Fields Code.”

Both these approaches rely on ensuring that each processing step returns complete data along with the linked metadata so subsequent nodes (like Ai Resume Fields) can access the necessary fields via item linking.

Feel free to ask if you need more details on a specific part of the implementation.