node package manager
Love JavaScript? Your insights can make it even better. Take the 2017 JavaScript Ecosystem Survey ยป



Build Status Code Climate

A tool to quickly populate your mongo db from a set of .json files. The concept is very similar to Ruby on Rails idea of seeding a database. This allows a new developer to pull down the source code for a project (which contains .json files to populate their mongodb with) and run a command and bingo!! database populated and ready to rock.



  • Run $ npm install -g node-mongo-seeds
  • Run $ seed-setup from the root of your project to generate a seed.json file
  • Replace "localhost/LOCAL_DB_NAME" with the path to your mongodb in your brand new seed.json file
  • Create a /seeds folder in your project root and put .json files in there. The name of the file is going to be the collection name in mongo and the contents of the file will be populated into that mongo collection.
  • Run $ seed to seed your mongodb with all your data from your /seeds folder.

Note: Every time you run $ seed it will blow away all the data in your collections and re-populate them with whatever is in your /seeds directory.

Separate databases per NODE_ENV

You can have separate databases for each NODE_ENV that you are using. Just a put a line in seed.json for each NODE_ENV that you are going to use. It defaults to undefined, dev, and prod but can be changed to whatever.


  1. Clone project and run npm install -g ./ from project root
  2. Add feature(s)
  3. Add tests for it
  4. Submit pull request

Running Tests

To run the tests, follow these instructions.

  1. Start mongod process
  2. From project root $ npm install
  3. Run test command $ npm test