Improve deployer error messages

[Siegrift]

See: https://api3workspace.slack.com/archives/C02AYRX8D89/p1663092715835849

Currently, the deployer spits out terraform error messages which are too low level for end users. Also, due to an implementation limitation, the error is printed twice.

What I propose is to:

• capture the terraform output and in case of an error provide just a short message
  • we can know contextually what operations is performed (update/redeploy/delete) and we can also parse the error message for common errors in the future (already existing gateway keys). For now use only a short generic error message.
• write full deployer debug information and error details into a file. The users can open this file to see more information about the problem and share the file to API3 to help debug. (The problematic part is that the file might contain sensitive information which should not be shared, e.g. stage, airnode address...).

This task could probably be split and could deserve a planning discussion on its own.