The subject is not well represented by Agamben. Almost everything about this topic is shrouded in the mists of time, poorly attested, and extremely debateable.

Notice a few things here:

• Agamben looks to a later source, Festus, for the idea, and claims that that's where 'the character of sacredness is tied for the first time to a human life as such' -- a good 200 years later than Dionysius.
• As you've noticed, he doesn't cite sources when outlining the crimes.
• Even when he cites Festus, he only cites the title: good luck tracking down the exact passage.
• Look at the modern authors he cites, and who he doesn't cite. He cites Durkheim, a sociologist; Foucault, a philosopher; Kerényi, a fringe scholar; Otto, a theologian.

He doesn't cite a single scholar of Roman law -- and this is in a dedicated discussion of Roman legal concepts. The only relevant historical scholarship he cites is an article from 1911. All of the problems he identifies in the meaning of the word 'sacred' are copied from that article, without reference to anything else.

Tl;dr: Agamben's entire knowledge of Roman law comes from a single article published in 1911.

Now, there are real things that his discussion is related to. The story he reports is the result of overinterpreting a very late, very inaccurate report of something from mythical times.

I refer to an 8th century CE set of excerpts, made by Paul the Deacon (Paulus Diaconus) from another abridgement, of Festus' 2nd century On the meaning of words. Paulus makes the following claim about Numa, one of the early mythical kings of Rome (p. 505 lines 19-21 ed. Lindsay; Paulus' excerpts aren't present in Müller's edition of Festus, which is the text in the PHI database):

Termino sacra faciebant, quod in eius tutela fines agrorum esse putabant. Denique Numa Pompilius statuit, eum, qui terminum exarasset, et ipsum et boves sacros esse.

People used to make things sacred to Terminus, because they thought land boundaries were under his protection. In the end Numa Pompilius decreed that if anyone ploughed up a terminus, he and his oxen were sacred.

'Sacred' in this passage is regularly interpreted in reference to a Roman practice of sacratio or consecratio, devoting property or people to a god, making them the property of that god, and/or forfeiting them to the god for destruction. In this sense, a homo sacer was an outlaw.

Sacratio is well attested elsewhere -- just not in connection with moving boundary markers. We do have reports of laws in early Rome whereby a person who assaulted a tribune could be killed without penalty, attested by Dionysius of Halicarnassus 6.89.3, referring to a law of 494 BCE; and that someone who declared the election of a magistrate without allowing an appeal could suffer the same penalty, attested by Livy 3.55.4-7 in reference to a law of 449 BCE.

But things are absolutely not clear. All these stories are extremely early, dealing with practices, laws, and people who are as much myth as anything else. And you'll notice that the Dionysius and Livy passages don't use the key word 'sacred'.

It's often inferred that Festus' story is about moving boundary markers of private property: that's the interpretation of Geoffrey Mac Cormack in this 1979 article, for example, informed by a passage in Dionysius of Halicarnassus about Numa ordering people to place boundary markers on their land.

But sacredness of boundaries is much more tied up with city boundaries, such as the pomerium, the religious boundary of Rome on the Palatine hill, beyond which auspices could not be taken; or the boundary between Rome and neighbouring cities. As far as I'm aware, situations where sacratio is applied to private disputes are about consecrating an offender's property, not their person (e.g. Cicero, De domo 123-125). I think public boundaries are a much more likely interpretation, supposing that Paulus' story holds any water at all. Consider an archaic stone in the Comitium, apparently dating to the monarchic period, CIL 1² 2.1:

quoi hoi . . . sakros esed sor . . . ia . ias recei ic . . . evam quos re . . . m kalatorem ha . . . iod iouxmenta kapia dotav . . .

In which some words can be straightforwardly recognised: here they are put into classical spelling:

qui hoc . . . sacrus esto . . . regi . . . iumenta . . .

He who ... let him be sacred ... to the king ... oxen ...

It's very tempting to connect this to Paulus' story (as does Louise Adams Holland, in this 1933 article). Something along comparable lines to: 'He who moves this stone, let him be sacred, and he and his oxen be forfeited to the king'.

But it's very archaic. And that means we don't really know what was going on; we certainly don't know how standard it might ever have been in the prehistoric period to consecrate a person as a penalty. A person devoting themselves, as a pious religious act, is something we do hear about occasionally.

But by the time we get to the historical era, no one's talking about state-sanctioned murder. In the 1st century BCE Lex Iulia agraria ('Julian/Mamilian law on property'), the penalties for moving boundaries are purely monetary. The only other occurrence of the phrase terminos exarare is even later, in the 6th century CE Digest of Justinian (10.1.4.4), which clearly specifies judicial recourse rather than religious consecration.

If land markers (termini) are claimed to have been thrown down or ploughed up (exarati), the judge who is adjudicating the dispute has the power to make decisions about boundaries.

For the record, even public boundaries could be moved, and were moved, at times: the pomerium was moved at various points in Roman history, including by Sulla, Caesar, and Claudius (Aulus Gellius 13.14.4-7; Tacitus Annals 12.23).