Is there any specific set of instructions on how to Romanize Hebrew?

The Square Hebrew / SimHebrew™ Converter will convert any Square Hebrew text into Roman characters (and back again), with full fidelity with regard to spelling and distinction between seemingly homophonous characters:

As with French, German, Italian, or any other Roman-based language, however, you would have to bear in mind that certain characters have different phonetic values compared with English:

• a = alef, a vowel carrier, in absence of any supporting consonant (e.g., ama = /ima/; aocl = /okhel/)
• b = bet, which is hard at the beginning of words, ‘soft’ (/v/-like) at the end of words and sometimes in the middle as well
• k = ḥet, pronounced like the <j> in Spanish or <ch> in loch
• i = yod, which is /y/-sounding at the beginning of words or syllables, /i/ sounding (as in pizza) on all other occasions
• c = kaf, which is hard (/k/) at the beginning of words, ‘soft’ (ḥet-like) at the end of words and sometimes in the middle as well
• ç = kaf sophit, the end-of-word form of kaf, which is always ‘soft’.
• y = ayin, a vowel carrier like aleph, which in theory should be pronounced in the throat, as in Arabic, but in practice is not, by most Israelis
• p = péh, which is hard at the beginning of words, ‘soft’ (/f/-like) at the end of words and sometimes in the middle as well
• x = tzadi, pronounced /tz/
• w = shin, which is usually pronounced /sh/, but occasionally /s/ (as in iwral = yisrael = Israel

The full mapping scheme (if you wish to convert by hand) is as follows:

As for knowing how the vowels go, the /o/ and /u/ are indicated if the vav is dotted, the i indicates an explicit /i/ sound, and other vowels are /a/ or /e/ that you either recognise or infer from the word’s context or conjugation.