(Divitias alius fulvo sibi congerat auro
Et teneat culti iugera multa soli,)
Quḗm lăbŏr ā́dsĭdŭū́s vīcī́nō tḗrrĕăt hṓste
Mā́rtĭă cū́i sōmnṓs clā́ssĭcă pū́lsă fŭgént
Rather than immediately propose a translation of the two verses, I will show my reasoning to allow someone perhaps to learn from or criticize my process.
First, we need to understand the context, which requires an understanding of the first two verses of the poem. I understand them in this way:
"As for riches, let another heap up them for yellow gold,
And let him hold many acres of tilled earth,"
From a discourse perspective, we should expect that any immediately following verses that approximate the same structure will address the same thought, amplify it, or expand on it from a different perspective.
I see two subclauses, one main clause but only two verbs. Two
subclauses are subordinated by 'quem' and 'cui' and the two verbs are
'terreat' and 'fugent', where is the missing verb?
Since the verses in questions begin with what appears to be a relative pronoun, we should keep in mind that Latin relative pronouns used near the beginning of a sentence or clause are often used when in English we would use "and" plus a pronoun, so that quem might be the equivalent of et eum/sē ("and him....") and cui might be the equivalent of et eī/sibī ("and to him/her/it....)"
For the first line of the question (the third of the poem), if we try to understand the words in order, we can assume that quem refers back to alius ("another") from the first line of the poem. It must be masculine accusative singular and animate and most probably is the object or target of a following verb. Since it is far removed from its antecedent, we should tentatively assume that it is indeed the equivalent of et eum/sē ("and him...")
The word labor is masculine nominative singular and likely the subject of a following action verb, since we already have an object. Although it can be translated with the English cognate "labor," it often has a more negative connotation in Latin veering into "toil," "hardship," and "trouble." Since so much of Latin syntax is semantically driven, we should tentatively use a stronger word, like "toil," in our mental translation until we know the full context.
The word adsiduus ("constant"/"unremitting") is masculine singular nominative and so must refer to labor attributively or predicatively, a question to be resolved by the semantics of the verb. Since it is used with a noun with negative connotations, a tentative translation like "unremitting" is in order, giving us "unremitting toil/trouble" or "toil/trouble doing something unremittingly."
The word vīcīnō ("nearby"/"neighboring") is masculine or neuter dative or ablative singular and is tentatively semantically incompatible with anything so far and is in a prime position before a verb for hyperbaton or scrambling, probably indicating that it is emphasized. It may be attributive, but, particularly if it is ablative, it may be predicative. We have to wait for the verb to understand its specific case and semantics.
The word terreat is a verb parallel to teneat in the second verse of the poem so is presumed to continue its semantics, which here have an injunctive force, i.e., "let X terrify Y." We already know its two arguments and that X must be labor and Y must be quem, giving us a translation so far like "and let (unremitting) toil/trouble terrify him (without cease)." The semantics of terreat make a dative very unlikely, so vīcīnō is likely some kind of ablative, giving us something like "and let toil/trouble terrify him unrelentingly with/in a neighboring something or with/in something that is nearby."
With the last word in the verse, hoste ("foe"/"enemy"), we have all the arguments and all the adjuncts encountered so far taken care of. If we understand the pragmatics of the verse as upping the ante of the first two verses, then we can arrive at something like: "And let trouble terrify him without cease with an enemy nearby." Presumably, this is the expected consequence of having "heaped up riches" and "many acres of tilled soil": you now have to fear that someone nearby will want to take it from you.
With the last verse, we expect a thought continuing the previous verse or summarizing everything so far.
The fourth verse of the poem starts with Mārtia, which is feminine singular and nominative, given the meter, or else neuter nominative or accusative plural. Since nothing so far fits these specifications, we assume it is looking forward to something later in the sentence, and we understand that something or some things martial/warlike do something or are in some state.
The word cui, as another form of the masculine relative pronoun, seems presumptively parallel to quem since it is also near the beginning of its verse and clause. We should also assume that cui and quem have the same referent, unless something else indicates otherwise. If so, the whole verse should be repeating the same type of thought as the verse/clause with quem, and we should be looking for other consequences of having riches and much land. Syntactically, this should also mean that cui has been postponed after Mārtia for the meter or for effect, leaving us with "and a Martial something or somethings will be in some relationship to him."
The word somnōs ("sleep"(s)) is definitely masculine plural accusative. Latin poetry often uses plural for singular regardless of the normal usage of this noun, so we can assume it simply translates "sleep" or "slumber." We now can be sure that Mārtia is nominative, with the accusative slot already filled by somnōs. Since "sleep" is arguably an inalienable noun, we can also suspect that the relationship expressed by the dative of cui would be translated by an English possessive pronoun, giving us "something warlike or some warlike things are doing X to his slumber."
The word classica could mean many things, but the ending forces us to link it with Mārtia, giving the only likely meaning as "trumpet calls" in the neuter nominative plural. We now have "Warlike trumpets are doing X to his sleep" and can already guess at the semantics of the verb.
The word pulsa ("struck") presumptively agrees with Mārtia and classica and so is probably neuter nominative plural. Pellō can mean to have a musical instrument make its characteristic sound and so fits with the semantics of "trumpet." We can translate it as "sounded" to suit English semantics for trumpets. Since it is late in its clause, we can look for a predicative meaning, giving us "Warlike trumpets are doing X to his sleep as they are sounded."
The last word is fugent ("cause to flee") and is parallel in mood to all the injunctive verbs before. Since it is plural, it agrees with Mārtia classica. Now we have "Let warlike trumpets set his sleep to flight as they sound," again expressing some of the worries that will accompany the possession of "heaped riches" and "many acres of tilled earth."
To close, I should clarify that all the injunctive verbs need not be taken as literal commands or even the wishes of the poet. The injunctive sense is wide enough to capture that the poet may be proposing hypotheticals for the reader to ponder. A different idiom for this could be captured by English imperatives: e.g., "Have riches, and have troubles."