Add OpenAI content policy API checks

[headchem]

Requirements: https://beta.openai.com/docs/usage-guidelines/default-safety-requirements-for-all-applications

• Implement the free OpenAI Content Filter to prevent Unsafe (CF=2) completions from being returned to the end-user.
• Limit the max_tokens parameter to 250 per completion.
• Send no more than 1000 characters of user input per request. (toward end of the story, we may be including more than this amount of user-generating Sequence descriptions and Character descriptions)
• Implement end-user rate limits of no more than 60 requests per minute
• To help with monitoring for possible misuse, developers serving multiple end-users should pass an additional user parameter to OpenAI with each API call, in which user is a unique ID representing a particular end-user. Example:

response = openai.Completion.create(
  engine="davinci",
  prompt="This is a test",
  max_tokens=5,
  user="1"
)

[headchem]

I can also use the content filter API to periodically scan existing stories for inappropriate content, which could automatically prevent public sharing? Or to alert me to block an account.

Make a new Service layer dedicated to content checking so I can easily use it for both completions and general scanning.