Maybe prokka?

Prokka, like the more recent Bakta (never tried the latter so it might be different in the details), are suites of programs for genome annotation: gene calling + assigning function. Quite far from what you ask for and, AFIAK, one can't skip the gene calling step with one's own ORFs, so no way for protein sequences.

I will be very sad if I can't multiqc in my pipeline anywhere at all xD

I'd go for custom like a summary of how many clusters were found with how many proteins etc.

Yep, that is definitely interesting. Cluster size distribution plots as well. Per family sequence length distributions