Querying

Mongster queries stay close to MongoDB. The main difference is that projections, populate, sorting, and a few other helpers stay typed.

Read queries

import { Todo } from "./models";

const ownerId = new ObjectId();

const query = Todo.find({ ownerId, completed: false })
  .include(["title", "tags", "createdAt"])
  .sort({ createdAt: -1 })
  .skip(10)
  .limit(20);

const todos = await query;

find(), findOne(), and findById() return thenable query builders. You can await them directly, but you can also keep building the query before it runs.

Updates and counts

import { Todo } from "./models";

await Todo.updateOne({ _id: todoId }, { $set: { completed: true }, $push: { tags: "done" } });

const updated = await Todo.findById(todoId);
const total = await Todo.count({ completed: true });
const owners = await Todo.distinct("ownerId", { completed: true });

Common helpers

• include() and exclude() for typed projections.
• project() when you want to pass a projection record yourself.
• sort(), skip(), and limit() for cursor shaping.
• count(), estimatedCount(), and distinct() for collection summaries.
• explain() returns MongoDB's query plan for the same filter and options.
• getCursor() hands you the raw FindCursor for streaming or advanced workflows.

Updates, counts, and operators

For the rules Mongster applies to update operators ($set, $inc, $push, $addToSet, $rename, ...), see the Updates guide. That page lists every supported operator with an example and the validation it triggers.

Populate has its own guide because refs come with a few important limitations.