trello-track

Track running processes on Trello.

Why?

Let's say you're running experiments on several machines. Wouldn't it be nice to have an easy way to see when everything is done and if anything has encountered an error?

trello-track addresses this issue by adding checklist items to a Trello card that are updated to reflect the progress of a process. You can quickly see at a glance where all your experiments are at!

Features:

• Command line and python interfaces
• Task status indicators:
  • ⚪ ready
  • ⌛ in progress
  • 🔵 success
  • 🔴 failed
• Error logging

How?

There's two interfaces: a command line interface and a Python interface.

They both take the Trello Card ID. This is easy to pluck out of the URL.

CLI

Format: trello-track [card-id] [command...]

Example: trello-track 9ngUrIZ2 bash train_eval.sh A X 0.1

Python

import trello_track

with trello_track.track('do something', card_id='9ngUrIZ2'):
	foo()

But you probably don't want to hard-code a card ID, do you? You can instead skip the optional parameter and provide it via an environment variable:

TRELLO_CARD=9ngUrIZ2 python ...

You can also queue up a bunch of tasks to run using trello_track.TaskManager:

import trello_track

with trello_track.TaskManager() as tasks:
	tasks.add('foo', foo)
	tasks.add('bar', bar)
	tasks.add('baz', baz)

Authentication

Authentication is done via the Trello API key and token.

You can access your key and generate a token at https://trello.com/app-key.

You then can provide them to your application via the TRELLO_KEY and TRELLO_TOKEN environment variables, or encoded as json in ~/.trello or ./.trello:

{"key": "XXX", "token": "XXX"}