The stated question:
The problem with the situation you describe is that the two characters were operating in a non-combat context, where issues like Initiative, turns, and simultaneous activity in a span of time don't apply. There is no way to ready an Action in a context where Actions and Reactions aren't operative.
A DM can try to shoehorn in some mechanic-esque elements, such as allowing the Cleric to roll an Insight check to see if they could have predicted that the Warlock was going to blast the werewolf in time to intercede, but that can quickly become awkward-- every action one PC takes that another would want to head off would be subject to mild retconning, or time and turn order would have to be tracked all the time, which is tedious.
D&D 5e was not built for this sort of PvP action in such granularity outside of combat. As a result it's asking a lot of the game system to adjudicate these situations. And, frankly, I'm not sure that a mechanical approach to deciding who "wins" all the time would do much to address the underlying problems.
If your table wants such a mechanic anyways, it'll have to be homebrewed, and designing it will require some specific goals and priorities be determined beyond plain mechanics.
There are two broad ways to think about this dynamic: as the DM's problem, and as the players' problem. Each has some different features, and which track you follow is a matter of preference.
This is the DM's problem:
Weave the conflict into the plot explicitly.
In my experience this kind of inter-PC dissention is at its worst when it comes to spontaneous plans, exactly like what was described in the question. The PCs can agree (at least tentatively) to work together when fighting a boss that threatens both of them, and players are usually knowledgeable enough about game tropes to appreciate when their interests align in that way, plot-wise, and when they don't (such as a side-quest with no bearing on the main plot).
This is a problem because the players are trying to get into their roleplaying at moments which seem convenient and appropriate, but they can't both do so at once. Without any structure to their conflicting approaches it's not clear when, or why, one PC would stand down in favor of the other, or why one would bother to try changing their methods.
It is the DM's explicit responsibility to keep players on track and keep the game moving, and they have a lot of tools to use to do it. Blunt options usually involve plot elements intended to nudge players towards certain behaviors, or which at least cause players to take some estimable risk in doing things other ways:
- If the players interact with a deity or powerful/influential figure
in an area, that figure might explicitly warn them against actions
like murder (and have the resources to know if they commit murder, and back up the prohibition).
- Avoid throwaway NPCs, so that consequences are expected. If the
werewolf were also a nobleman, the death might be a much bigger
deal to the story than a random enemy
- Enforce alignment. Not in the sense of "you're N/N, so you can't do
X!", but more like "this campaign is written for non-evil characters,
and this NPC or organization won't interact with an evil character".
It's not impossible, but definitely a stretch, for a Neutral/Neutral
character to kill a lone creature that is not threatening them and
has guaranteed it will not do so as it walks away. Alignments aren't straightjackets, but they are a good plot tool
- Use guilt-free mooks. If the werewolf were a member of some group
committed to killing the party (or innocent people, or whatever) then
the C/G Cleric's aim of rescuing and freeing it might not be feasible
- Mix it up. If there are some situations where a "good" option is
desirable, and difficult or impossible to elide, there should be some
situations on the other side where the Warlock's preferred approaches
are acceptable or desirable. No one should get their own way all the
time, but neither should they get their way none of the time
- Enforce consequences in every direction. When the outcome doesn't
strictly matter, then this is only RP flavor. That's often fine, but
in cases like the one you describe that just means that neither PC
has a better argument than the other for why their actions were
justified. If there are always consequences that bud off of actions
where there is some dispute, then the DM is offering some structure
and order that the players can use to make decisions in-game, which
will at least get them to discuss things on common ground. "If we do
X, we might have to deal with Y" has a conclusion which impacts both
characters equally, while "I want to do X because I'm Y vs. I want to
do Z because I'm A" doesn't necessarily have a common resolution.
If these aren't enough, the conflict between the PCs can be made a stronger plot element. Perhaps the main quest-giver interacting with the party is the C/G Cleric's deity, and that deity has the power to demand that the Warlock behave in a certain way while carrying out the deity's orders. Importantly, the deity can impose some sort of consequences for non-compliance.
If the DM is feeling expansive, contention between organizations directing each PC (or just aligning with their beliefs and values) can be a major point of contention in the plot, where empowering one group involves weakening the other. The players will have to agree on a way forward, whichever group they follow. While the plot can be resolved either way, there may be substantial changes in the content and difficulty of the campaign as it progresses based on the choices they make. This is the most common "invisible" way I handle these kinds of disputes in-game.
In any case, an out-of-character discussion with the relevant players and DM is probably the best way to get everyone on the same page. It's difficult to involve the DM in an in-game way.
This isn't the DM's problem.
The DM controls everything in the game except PC behaviors and interactions. I'm open to low-key ways of keeping PC groups together, as in the section above, but I also make it clear to my players that I am not interested in forcing them to stick together. The party should supply its own reasons for existing and remaining a party, and if they can't assemble a rationale then maybe some or all of the players should make some adjustments.
At the outset of games I expressly indicate that I don't want to invest much time or effort into taking this task on. If I must do so, I will address this problem once, at most. Any solution I come up with and impose may or may not be satisfying to any or all players, but will be definitive and may involve elements from strong railroading to substantial punishments for falling out of line. The idea is that I can't force fun for anyone, but I can force order which allows the game to continue.
So if players can't find a way to respect each other's fun in allowing them to play their characters as they want but still stay together, then maybe one PC should be put on a shelf for another game and a different character be rolled. Stagnating in conflict indefinitely only enhances my motivation to remove the conflict, by any means, which can have unpredictable (to the players) consequences on the game they're playing. If that means that the setting changes from the Material Plane to the Abyss or a celestial realm, where certain types of action are circumscribed or extremely risky, so be it. As DM I'm there to run a game that everyone plays, not to force everyone to play a game they don't want to continue.