Right now, if we add a facet stage into the aggregation pipeline, we have to write the rest of the aggregation query manually. Users should be able to edit the sub-pipelines within a facet stage.

When we create a pipeline, it'd help if we could check the same documents data in the different stages ( now their preview is from randomly selected documents AFAIK ), so it's inconvenient to inspect one specific documents preview phases through all the stages.
Possible solutions:
1) lock the first stage's documents' order and apply it to the following stages (if possible, eg. if the stage is $group, then we can not apply the rule to the next stage)
2) set a sorting ( if possible ) of documents. But only for the preview.

The pipeline builder is powerful for constructing an aggregation pipeline. However, it would be good to copy an entire pre-written pipeline into the aggregation builder as text and have it syntactically divide the text into the aggregation builder stages UI.

This will help make pipelines written previously and not saved on Compass easily modifiable over copying each stage manually into the builder UI, which is a tedious process for long pipelines.

It's a human tendency to forget to continuously save an authored pipeline while constructing it. We might create a large pipeline and forget to save it before closing MongoDB Compass.

Similar, to other softwares please add a prompt to Compass in a situation when changes are not saved before closing it.

I know you can import the pipeline from a language but I want to completely disable GUI pipeline builder and just have a text editor. The UI is too slow and clunky, just give me a text editor to directly edit and run an aggregate.

The intent would be to run SQL queries directly on MongoDB Compass or convert them to aggregations

It would be great, if there would be an option to resize the input-fields within Aggregation-Editor

When I have "Builders" selected, all I get for an error is "$eq cannot be top-level". Does not help the user at all to know what this means.

I would like to be able to export pipelines in R syntax, as I can today in Node, Java and others. R is very picky and unforgiving about it's JSON so having it generated automatically would save a lot of time.

In aggregation view, add the possibility to add a stage Before or After.

MongoDB Compass + Atlas

The current Aggregation Pipeline Builder is good for authoring pipelines, but not so much for running them. The final results are limited to 20 docs and not easy to view, and viewing the intermediate stage outputs can slow down performance. There should be a way to execute a pipeline and page through the complete results.

Aggregations can be saved in Compass but they are only accessible in the Aggregation Builder for the specific cluster and namespace. It would be good to be able to reuse pipelines across different clusters and namespaces (e.g. I build it against my staging database but then I want to also test it in prod).

It would also be great to have a way to share a library of pipelines with the rest of the team. In the ideal scenario, pipelines would be versioned and multiple people would be able to contribute to a shared library of pipelines.