feat(compass-collection): Create Mock Data Generator Prompt and Batching Logic in Compass - CLOUDP-381914 by jcobis · Pull Request #7892 · mongodb-js/compass

jcobis · 2026-03-19T21:48:29Z

Description

Create packages/compass-generative-ai/src/mock-data-generator/ module with Zod tool schema, system prompt, user prompt formatter, and schema batching logic, all ported from the MMS Java backend and translated into TypeScript
Move MockDataSchemaResponseShape, MockDataSchemaRawField, and MockDataSchemaResponse from atlas-ai-service.ts into the new module (re-exported for backwards compatibility until next ticket switches over)
Add unit tests for batching logic (covering splitSchemaIntoChunks, mergeChunkResponses, validateSchemaSize, needsBatching)

Checklist

New tests and/or benchmarks are included
Documentation is changed or added
If this change updates the UI, screenshots/videos are added and a design review is requested
If this change could impact the load on the MongoDB cluster, please describe the expected and worst case impact
I have signed the MongoDB Contributor License Agreement (https://www.mongodb.com/legal/contributor-agreement)

Motivation and Context

Part of the broader effort to migrate all Atlas AI calls to the EDU Knowledge Server.

Bugfix
New feature
Dependency update
Misc

Types of changes

Backport Needed
Patch (non-breaking change which fixes an issue)
Minor (non-breaking change which adds functionality)
Major (fix or feature that would cause existing functionality to change)

jcobis · 2026-03-19T21:50:44Z

packages/compass-generative-ai/src/mock-data-generator/format-schema-for-prompt.ts

Ported from mms: MockDataSchemaGenerationPrompt.buildUserPrompt()

jcobis · 2026-03-19T21:54:52Z

packages/compass-generative-ai/src/mock-data-generator/prompt.ts

Ported from mms: MockDataSchemaGenerationPrompt.java

jcobis · 2026-03-19T21:55:22Z

packages/compass-generative-ai/src/mock-data-generator/schema-batching.ts

+ * Splits a schema into smaller chunks for processing.
+ * Ported from NaturalLanguageQueryGenerator.splitSchemaIntoChunks()
+ */
+export function splitSchemaIntoChunks(


Ported from mms: NaturalLanguageQueryGenerator.splitSchemaIntoChunks()

jcobis · 2026-03-19T21:55:36Z

packages/compass-generative-ai/src/mock-data-generator/schema-batching.ts

+ * Merges multiple chunk responses into a single response.
+ * Ported from NaturalLanguageQueryGenerator.mergeChunkResponses()
+ */
+export function mergeChunkResponses(


Ported from mms: NaturalLanguageQueryGenerator.mergeChunkResponses()

Copilot

Pull request overview

This PR introduces a new mock-data-generator module in compass-generative-ai to support generating mock data schema mappings (faker.js field mappings) with prompt content, tool schema definitions, and schema batching utilities, while keeping existing exports working via re-exports.

Changes:

Added Zod tool schema/types and prompt text for mock-data-schema generation in a new mock-data-generator/ module.
Implemented schema batching utilities (split/merge/size validation) and added unit tests for them.
Updated atlas-ai-service.ts to consume the moved schema types/shapes from the new module and re-export them for compatibility.

Reviewed changes

Copilot reviewed 7 out of 7 changed files in this pull request and generated 3 comments.

Show a summary per file

File	Description
packages/compass-generative-ai/src/mock-data-generator/schema.ts	Adds Zod tool schema + moved response types with compatibility aliases.
packages/compass-generative-ai/src/mock-data-generator/schema-batching.ts	Adds chunking/merging/limits helpers for large schemas.
packages/compass-generative-ai/src/mock-data-generator/schema-batching.spec.ts	Adds unit tests for batching utilities.
packages/compass-generative-ai/src/mock-data-generator/prompt.ts	Adds the system prompt used to generate faker mappings.
packages/compass-generative-ai/src/mock-data-generator/format-schema-for-prompt.ts	Adds user-prompt formatting for schema + validation rules.
packages/compass-generative-ai/src/mock-data-generator/index.ts	Exposes the new module’s public surface area.
packages/compass-generative-ai/src/atlas-ai-service.ts	Switches mock-data schema response shape/type to import from the new module and re-exports for compatibility.

packages/compass-generative-ai/src/mock-data-generator/format-schema-for-prompt.ts

packages/compass-generative-ai/src/mock-data-generator/schema-batching.ts

packages/compass-generative-ai/src/mock-data-generator/schema-batching.spec.ts

mabaasit · 2026-03-30T10:15:13Z

packages/compass-generative-ai/src/mock-data-generator/schema-batching.ts

+  if (fieldsPerChunk <= 0) {
+    throw new Error('fieldsPerChunk must be a positive integer');
+  }
+


We can use lodash chunk to clean this:

return chunk(Object.entries(rawSchema), fieldsPerChunk).map( (chunk) => Object.fromEntries(chunk) );

mabaasit · 2026-03-30T10:35:42Z

packages/compass-generative-ai/src/mock-data-generator/format-schema-for-prompt.ts

+
+${schemaJson}
+
+${validationRulesPhrase}


Do we want to add validation rules to the prompt example as well?

Original prompt didn't include this, so how about we keep it the same for now? We tested the original prompt for accuracy and such, so don't want to include too many changes at once. I will implement your other suggestions though

mabaasit · 2026-03-30T10:37:52Z

packages/compass-generative-ai/src/mock-data-generator/format-schema-for-prompt.ts

+  documentSchema: RawSchema,
+  validationRules?: Record<string, unknown> | null
+): string {
+  const schemaJson = JSON.stringify(documentSchema, null, 2);


You can use toJSString function from mongodb-query-parser to achieve this (check utils/gen-ai-prompt).

mabaasit · 2026-03-30T10:39:17Z

packages/compass-generative-ai/src/mock-data-generator/format-schema-for-prompt.ts

+
+Documents in the collection are described by the following schema:
+
+${schemaJson}


Maybe wrap this in code fence (so that it is wrapped in backticks)

WIP

74d3f99

jcobis commented Mar 19, 2026

View reviewed changes

jcobis and others added 2 commits March 19, 2026 17:57

WIP

433a717

Merge branch 'main' into CLOUDP-381914

be06d03

jcobis changed the title ~~WIP~~ feat(compass-collection): Create Mock Data Generator Prompt and Batching Logic in Compass - CLOUDP-381914 Mar 24, 2026

github-actions bot added the feat label Mar 24, 2026

Merge branch 'main' into CLOUDP-381914

4e87be7

jcobis marked this pull request as ready for review March 24, 2026 17:43

jcobis requested a review from a team as a code owner March 24, 2026 17:43

jcobis requested review from Copilot and ivandevp March 24, 2026 17:43

Copilot started reviewing on behalf of jcobis March 24, 2026 17:44 View session

Undefined check

44ee528

Copilot AI reviewed Mar 24, 2026

View reviewed changes

Copilot comments

e28b767

jcobis requested a review from mabaasit March 24, 2026 17:57

jcobis added the no release notes Fix or feature not for release notes label Mar 24, 2026

mabaasit reviewed Mar 30, 2026

View reviewed changes

jcobis and others added 2 commits March 30, 2026 13:50

Address comments

d86f7f6

Merge branch 'main' into CLOUDP-381914

c00693f

jcobis requested a review from mabaasit March 30, 2026 17:51

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

feat(compass-collection): Create Mock Data Generator Prompt and Batching Logic in Compass - CLOUDP-381914#7892

feat(compass-collection): Create Mock Data Generator Prompt and Batching Logic in Compass - CLOUDP-381914#7892
jcobis wants to merge 8 commits intomainfrom
CLOUDP-381914

jcobis commented Mar 19, 2026 •

edited

Loading

Uh oh!

jcobis Mar 19, 2026

Uh oh!

jcobis Mar 19, 2026

Uh oh!

jcobis Mar 19, 2026

Uh oh!

jcobis Mar 19, 2026

Uh oh!

Copilot AI left a comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

mabaasit Mar 30, 2026

Uh oh!

mabaasit Mar 30, 2026

Uh oh!

jcobis Mar 30, 2026

Uh oh!

mabaasit Mar 30, 2026

Uh oh!

mabaasit Mar 30, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants


		Documents in the collection are described by the following schema:

		${schemaJson}


		${schemaJson}

		${validationRulesPhrase}

Conversation

jcobis commented Mar 19, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Description

Checklist

Motivation and Context

Types of changes

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Reviewed changes

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

jcobis commented Mar 19, 2026 •

edited

Loading