If a credential has special characters in its ID (e.g. /), the credential may not work or render properly in some parts of the Web UI.

For example, when the credential is transformed into a link, the link is broken (404 Not Found) because the '/' characters in the ID are not transformed into '%2F'.

This appears to be because there is no standardised urlencoding of credential IDs in the credentials API, so whether urlencoding happens or not is up to the concrete credentials providers.

It just so happens that the default (local disk) credentials provider does this - I believe through calling `Util.rawEncode(id)` in the CredentialsWrapper class. But a provider which does not have this logic will produce broken credential links.

I would propose that this logic should be moved up to the Credentials API so that it is done in one place, in a standardised way that applies to all providers.