[Resource]: Claude Code Hook Comms (HCOM)

[hesreallyhim]

Display Name

Claude Code Hook Comms (HCOM)

Category

Hooks

Sub-Category

None

Primary Link

https://github.com/aannoo/claude-hook-comms

Secondary Link

https://pypi.org/project/hcom

Author Name

aannoo

Author Link

https://github.com/aannoo

License

MIT

Other License

No response

Description

Lightweight CLI tool for real-time communication between Claude Code sub agents using hooks. Enables multi-agent collaboration with @-mention targeting, live dashboard monitoring, and zero-dependency implementation. [NOTE: At the time of posting, this resource is a little unstable - I'm sharing it anyway, because I think it's incredibly promising and creative. I hope by the time you read this, it is production-ready.]

Additional Comments

No response

Submission Checklist

• I have checked that this resource hasn't already been submitted
• My resource provides genuine value to Claude Code users, and any risks are clearly stated
• All provided links are working and publicly accessible
• I am submitting only ONE resource in this issue
• I understand that low-quality or duplicate submissions may be rejected

[hesreallyhim]

@aannoo this is pretty clever, dude/person... 🙌 nice job! so, basically, you use the SubAgentStop hook to keep the sub agent "alive" in order to have a persistent communication stream? That's really clever, I'll have to try it out, but great thinking.

As I said to another contributor, it's your project, so this is just my feedback, but maybe consider a different name? IMO calling it "Hook Comms" and focusing on hooks is not as clear, or important, as what the actual functionality is, which is sub agent comms - the fact that it uses hooks is "an implementation detail" - to me, Hook Comms makes it sound like the hooks are communicating or something. What do you think?

[aannoo]

@hesreallyhim Thanks! Yes, the Stop hook maintains the polling loop while Claude waits.
About the name - I can see what you mean but Hook Comms indicates it's using hooks, which makes it different from other multi-agent tools. I think the implementation method is relevant since it means no dependencies or external infrastructure, just the built-in hook system, files, and a python script. I wouldn't say it's just subagent comms as there's other stuff you could use it for.
Hopefully it's not confusing I guess it's hook-based comms rather than hooks communicating with each other.

[hesreallyhim]

@aannoo That's fine about the name it was just a suggestion, of course it's your project. The only thing to do now is wait unfortunately. Hope to get to it soon.

[hesreallyhim]

@aannoo damn i really wish i had time to review this already but i haven't forgotten about it fyi (although you're free to join that other user's revolution if you feel like it).

[hesreallyhim]

@aannoo ok so this is really freaking cool. When I started it up, the "demo"(?) kind of went crazy, the instances just started talking to each other and just kept going and going, eventually somehow I intervened and started to work with them to complete tasks. I'm not sure if it's something to do with interaction with existing hooks, but this is really really cool, as I suspected. I'm gonna give it another try but I'm really interested in this project, would you say it's "stable"? basically there was a lot of chatter going on and it was hard for me to "take control" and steer them the way I expected, but I love the concept.

[hesreallyhim]

/approve