Build Status Code Climate

A Ruby gem to interact with the Harvest API v2.0 and forward.


Add this line to your application's Gemfile:

gem 'harvesting'

And then execute:

$ bundle

Or install it yourself as:

$ gem install harvesting


The API is documented here


In order to start using this gem you will need your personal token and an account id:


You can find these details over here: https://id.getharvest.com/developers

If you don't specify values for access_token or account_id, it will default to these environment variables:


That means that you could build a client like this:

# $ export HARVEST_ACCOUNT_ID=12345678
client = Harvesting::Client.new
# => #<Harvesting::Models::User:0x007ff8830658f0 @attributes={"id"=>2108614, "first_name"=>"Ernesto", "last_name"=>"Tagwerker", ... >

If you don't specify a valid combination of token and account id, your code will raise this error:

client = Harvesting::Client.new(access_token: "foo", account_id: "bar")
# Harvesting::AuthenticationError: {"error":"invalid_token","error_description":"The access token provided is expired, revoked, malformed or invalid for other reasons."}

If your personal token and account id are valid, you should see something like this:

client = Harvesting::Client.new(access_token: "<your token here>", account_id: "<your account id here>")
user = client.me
# => #<Harvesting::Models::User:0x007ff8830658f0 @attributes={"id"=>2108614, "first_name"=>"Ernesto", "last_name"=>"Tagwerker", ... >

# => 2108614


# => #<Harvesting::Models::Clients:0x007ff718d65fd0>

client = client.clients.first
# => #<Harvesting::Models::Client:0x007ff718cf5fc8 @attributes={"id"=>6760580, "name"=>"Toto",
# ... >

Time Entries

time_entries = client.time_entries
# => #<Harvesting::Models::TimeEntries:0x007ff71913e3a0 @attributes={"per_page"=>100, "total_pages"=>1, "total_entries"=>14, "next_page"=>nil, "previous_page"=>nil, "page"=>1, "links"=>{"first"=>"https://api.harvestapp.com/v2/time_entries?limit=1&page=1&per_page=100", "next"=>nil, "previous"=>nil, "last"=>"https://api.harvestapp.com/v2/time_entries?limit=1&page=1&per_page=100"}}, ... >

entry = time_entries.first
# => #<Harvesting::Models::TimeEntry:0x007ff71913dfe0 @attributes={"id"=>792860513, "spent_date"=>"2018-05-14", "hours"=>1.0, "notes"=>"hacked the things", "is_locked"=>false, "locked_reason"=>nil, "is_closed"=>false, "is_billed"=>false, "timer_started_at"=>nil, "started_time"=>nil, "ended_time"=>nil, "is_running"=>false, "billable"=>true, "budgeted"=>false, "billable_rate"=>nil, "cost_rate ... >


tasks = client.tasks
# => #<Harvesting::Models::Tasks:0x007ff718897990 @attributes={"per_page"=>100, "total_pages"=>1, "total_entries"=>6, "next_page"=>nil, "previous_page"=>nil, "page"=>1, "links"=>{"first"=>"https://api.harvestapp.com/v2/tasks?page=1&per_page=100", "next"=>nil, ... >


projects = client.projects
# => #<Harvesting::Models::Projects:0x007ff718e1c8e8 @attributes={"per_page"=>100, "total_pages"=>1, "total_entries"=>1, "next_page"=>nil, "previous_page"=>nil, "page"=>1, "links"=>{"first"=>"https://api.harvestapp.com/v2/projects?page=1&per_page=100", ... >

project = projects.first
# => #<Harvesting::Models::Project:0x007ff718e1c618 @attributes={"id"=>17367712, "name"=>"Foo", "code"=>"", "is_active"=>true, "is_billable"=>true, "is_fixed_fee"=>false, "bill_by"=>"none", "budget"=>nil, "budget_by"=>"none", "budget_is_monthly"=>false, "notify_when_over_budget"=>false, "over_budget_notification_percentage"=>80.0, "show_budget_to_all"=>false, "created_at"=>"2018-05-13T03:30:06Z", ... >


invoices = client.invoices
# => #<Harvesting::Models::Invoices:0x00007fc8905671f0 @models={}, @attributes={"per_page"=>100, "total_pages"=>1, "total_entries"=>3, "next_page"=>nil, "previous_page"=>nil, "page"=>1, "links"=>{"first"=>"https://api.harvestapp.com/v2/invoices?page=1&per_page=100", ... >

invoice = invoices.first
# => #<Harvesting::Models::Invoice:0x00007f8c37eb6d18 @models={}, @attributes={"id"=>23831208, "client_key"=>"73688e97a43ed497ace45939eb76db6b18427b80", "number"=>"3", "purchase_order"=>"", "amount"=>750.0, "due_amount"=>750.0, "tax"=>nil, "tax_amount"=>0.0, "tax2"=>nil, "tax2_amount"=>0.0, "discount"=>nil, "discount_amount"=>0.0, "subject"=>"", "notes"=>"", "state"=>"draft", "period_start"=>nil, ... >

An invoice can have many line items:

line_items = invoice.line_items
# => [#<Harvesting::Models::LineItem:0x00007f92617ce8e0 @models={}, @attributes={"id"=>109677268, "kind"=>"Service", "description"=>"", "quantity"=>3.0, "unit_price"=>250.0, "amount"=>750.0, "taxed"=>false, "taxed2"=>false, "project"=>{"id"=>24566828, "name"=>"Harvest Billing Automation", "code"=>""}}, ... >]

You can filter invoices by various attributes. E.g. client.invoices(state: "draft") only returns invoices in a draft state.

Nested Attributes

The Harvest v2 API embeds some data in JSON objects. You can access nested attributes quite naturally. For example, to access the user id for a time entry instance, entry, use:


Or to access the name of the client on a project instance, project:



Deleting All Items

When you need to delete all items, care needs to be taken, because the API uses pagination. The following code will only delete data from every other page.

# WARNING - only deletes every other page
client.time_entries.each do |time_entry|

While iterating over items from the first page, all of those items will be deleted. This will result in moving items from the second page onto the first page. If there are only two pages, then the second page will be empty. If there are more than two pages, then the second page will now contain items which would have previously appeared on the third page. Deleting those items will move the items from the fourth page on to the third page, and so on.

Instead you need to make sure you get access to all of the time entry objects before you try to delete any of them. The easiest way to do this is to convert the Enumerable instance into an Array, by calling #to_a, before you iterate over it.

# GOOD - This should do what you want
client.time_entries.to_a.each do |time_entry|


There are many things to be developed for this gem. For now they are tracked here: TODO.md


You can find more info here


After checking out the repo, run bin/setup to install dependencies. Then, run rake spec to run the tests. You can also run bin/console for an interactive prompt that will allow you to experiment.


To install this gem onto your local machine, run bundle exec rake install.

Using Docker

This setup allows you to create a completely isolated development environment for working on this gem. The version of ruby used for development and the gems that it depends on will remain inside the container.

After checking out the repo, run docker-compose build to create the gem container. Running docker-compose run gem bash will get you a bash session inside the container. From there, you can follow the instructions in the "Without Docker" section above.

Running Tests

The tests in this project use the VCR gem to record and playback all interactions with the Harvest API. This allows you to run the test suite without having an account at Harvest for testing.

If you add a test that requires making an additional API call, then you'll need to make adjustments to the .env file to provide account details that are required by the test suite.

WARNING: The test suite is destructive. It deletes all entities from the account before it runs. DO NOT run it against a Harvest account which contains information that you need to preserve. Instead, create a Harvest account for testing.

If you need to refresh the VCR cassettes, the easiest way is to delete all of the files located under fixtures/vcr_cassettes. The next time the test suite is run, VCR will make actual calls against the Harvest API and record the responses into updated cassette files.

Effort has been taken to ensure that private information is excluded from the recorded cassettes. To adjust this further, add additional filter_sensitive_data calls to spec/spec_helper.rb.


The models in this project reflect the Harvest v2 API endpoints:

  • client
  • contact
  • task
  • task_assignment
  • project
  • user
  • time_entry

There are also models for the Harvest v2 API collection endpoints:

  • clients
  • contacts
  • tasks
  • task_assignments
  • projects
  • users
  • time_entries

These collection models handle the Harvest v2 API pagination automatically, making it easy to enumerate through all the instances of each type.

The models try to reduce code duplication through base class helper functions to automatically define accessors for the attributes included in each type returned by the Harvest v2 API. The Harvesting::Base::attributed method will define accessors for each simple attribute included in an array passed as an argument. Data is returned from these accessors as strings.

Some data is returned from Harvest as nested JSON (e.g. time_entry.project.name). A base class helper to expose this using the available models is also present. The Harvesting::Base::modeled method will define accessors for each object attribute included in options. Both the name of the attribute and the model to use in accessing that data is supplied. Data is returned from these accessors as model objects.

NOTE: Nesting model objects requires that the nested model types be defined before the nesting model type. E.g. if project contains a nested client, then client must be defined before project in the harvesting.rb include list.


To release a new version, update the version number in version.rb, and then run bundle exec rake release, which will create a git tag for the version, push git commits and tags, and push the .gem file to rubygems.org.

We follow semantic versioning for version numbers: https://semver.org


Bug reports and pull requests are welcome on GitHub at https://github.com/fastruby/harvesting. This project is intended to be a safe, welcoming space for collaboration, and contributors are expected to adhere to the Contributor Covenant code of conduct.


The gem is available as open source under the terms of the MIT License.

Code of Conduct

Everyone interacting in the Harvesting project’s codebases, issue trackers, chat rooms and mailing lists is expected to follow the code of conduct.