2lemetry ThingFabric Command Line Interface

pip install pytf==0.2



Because curling is for Canadians.

tf-cli provides a Python module (pytf) that includes a command line tool (tf) that wraps the 2lemetry ThingFabric API. At this point in time, both tf-cli and the ThingFabric APIv3 are both heavily in beta, so there should be plenty of caution, at least until tf-cli gets a 1.0 release.

The command line tool is heavily inspired by the the AWS CLI, so if you're familiar with that, this is an easy tool to pick up.


The easiest way to install is using pip:

$ sudo pip install pytf

If you'd like to install from the Github repository itself (in order to do some development, etc.), simply clone it, use virtualenv to set up the environment, and hack away.

$ git clone
$ cd tf-cli
$ virtualenv venv
$ source venv/bin/activate
(venv)$ python ./ develop

After you've installed pytf the first thing you should do is set up your ThingFabric credentials file. This will contain your access_key and your secret_key. Create a file in ~/.tf and fill it in with your credentials:

access_key = 01234567-89ab-cdef-0123-456789abcdef
secret_key = fedcba98-7654-3210-fedc-ba9876543210
url =

If you do not set your credentials here, the tool with attempt to read TF_ACCESS_KEY and TF_SECRET_KEY from the environment. If this fails, it'll throw a fit. Also note that you do not have to set the ThingFabric API URL here; if you omit it, it will just fall back to this URL. If you'd rather use the test URL or not use HTTPS (not recommended, dummies), then set it here.

Command Line Usage

The easiest way to see what tf can do is to just ask it:

$ tf help

This will generate a list of all of the available commands. These roughly map one-to-one to the ThingFabric APIv3. I took a few liberties to create a consistent naming convention and overloaded a function or two where I could. That said, it's not rocket surgery.

To take a look at what an invdiviual command can do, just ask it:

$ tf get-present-thing help

The arguments listed roughly map to one-to-one to the APIv3 arguments. Again, liberties were taken.

The command will result in the JSON response from the normal API call, and will exit 0 if the command succeded and exit 1 if not. There's also a cute option, --select, that's available on each command which allows you to select a particular item out of the JSON response.

$ tf nested-command-response
    "foo": [
            "bar": "rab",
            "baz": "zab"
$ tf nested-command-response --select
Module Usage

The CLI is generated directly from annotations in the module itself, so most of the documentation can be gathered from that. It's should be easy enough to understand, though documentation on PyPI will be coming soon(ish). Short of that the source code, REPL, dir, and inspect.getargspec are always your friends :)

from pytf import PyTF

api = PyTF(access_key, secret_key)
(result, ok) = api.create_token(ttl=3600)
token = result['authToken']