Audit Add-Ons/ Quality control checklist

[duckduckgrayduck]

Came to mind today that we have a large collection of Add-Ons now and occasionally some small issues arise here and there. It would be nice to do an audit. Here are some of the things that should be done for each Add-On for quality control.

• Does the Add-On have a category (AI, File, Extraction)?

• Is the Add-On listed here? https://www.documentcloud.org/help/add-ons

• Is python-documentcloud set to the most recent stable version of python-documentcloud (at the time of this issue creation 4.1.3)?

• Does the Add-On use self.get_documents() to iterate through documents for readability?

• If the Add-On offers the choice of running on a selection or query of documents, does it use the selected documents as default?

• Does the Add-On have a reasonable timeout or use SoftTimeOut if eligible?

• pylint/black?

• If the Add-On saves data back to DocumentCloud (save(), put(), delete(), etc), do we wrap API calls in try/except and handle the errors providing useful error messages to the user?

• If the Add-On makes network requests or outside API calls do we use exponential back-off with tenacity where appropriate?