Documentation thoroughness

[gpullela]

I was trying to setup zeroclaw on my computer and was able to build it from source with no issues.

My main issue was jumping between google, www.zeroclawlabs.ai/, the github docs, github issues, and discussions to figure out how to get my zeroclaw instance setup.

For example - I saw that a pull request regarding TOTP auth was made and then subsequently merged, but there is no documentation on how to set that up (that I could find atleast, might be woefully wrong, excuse me), if that is even possible.

I also had trouble figuring out how to use the GUI and setup connections with stuff like discord - thought I could do that by clicking on the discord integration (clicking on any of the integrations seems to do nothing but that's a discussion for later). I am not sure how to setup multi-user login (I am trying to convince my brother to use it so he can learn about agents and spritz up his discord server) and also how multi-agent setup works (if it is even possible?).

I have seen similar posts on youtube and other platforms that say zeroclaw seems great at first glance, but some of them seem to shy away because of the lack of documentation/setup guidance. Of course for those of you that manage to set it up, I'm sure once you figure it out it is fine.

Grass is much greener when you can see it once you stop pushing a boulder up the same hill.
-Sisyphus (probably)

I just personally wish the documentation was a little more extensive so that I could maybe go over (with my agent if I can get it setup). I have given up on openclaw (too much for my old 2017 gaming pc that struggles as it is), but I don't want to give up here. Anyone else feel like the documentation could be more extensive, and potentially more up-to-date/have release view maybe?

TLDR: I personally think the docs could be more detailed and maybe separated by version.

[arosboro]

Nothing serious until a valid CHANGELOG.md can demonstrate why we need so many departures from semver versioning standards

[gpullela]

You are seriously right - did not realize that there wasn't a concrete changelog in the repo until you mentioned it.

I got routed to this link by google. Something like this (openclaw's changelog.md) would definitely make tracking repo changes and potentially tracking required changes to the docs a whole lot easier.

Thanks again for pointing that out @arosboro :).

P.S. I personally feel like a feature parity document that tracks proposed/planned features would improve overall development and progress tracking, along with increasing transparency that I think would help increase adoption. Thoughts?

[arosboro]

Hi @gpullela, Gourav, I like that idea. Feature parity w/ proposed/planned mappings could be very helpful. Early on (v0.1.7 and prior) I had focused on developing a custom skill, connecting w/ a private matrix server, and readying for Standard Operating Procedures (SOPs which never landed). It's really discouraging seeing the level of whimsy at which features come and go. It's also difficult to plan areas for improvement on the agents I've tinkered with. I even went as far as writing a configuration manager to port my subagents and various config as the toml schema became a moving target.

Currently I've stopped w/ any attempt to run zeroclaw agents, however.

[gpullela]

Thank you for your kind words Andrew @arosboro :). I've also put running anything on hold for now as well. I would join the community call on discord if it wasn't at a weird time for me, but I think these are definitely things worth bringing up.

On another note, @theonlyhennygod @JordanTheJet is there any chance either of you plan to setup any polls post-discussion tomorrow? I feel like it could be a good way to gauge community consensus if you are open to it.

[JordanTheJet]

Hi there, sorry for getting back so late on this. Managing mountains of notifications. The last couple weeks I've been trying to lead the project to be more stable and align with standards. We made a milestones tab: https://github.com/zeroclaw-labs/zeroclaw/milestones and are trying to have more reasonable minor versions. Currently, we are not adding any features for the most part and are just refactoring the monolith into smaller parts for long term ability to follow the vision and improving the ability to develop on the project (docs, CI/CD, public skills). I think setting up polls are good. We also have contributor calls Wednesday nights and Open community roadmap calls Thursday nights. This version 0.7 is about the refactor and developer experience. The next minor version 0.8.0 will be about adding multi agent capabilities coordinated through the gateway/web UI which was in high demand in the roadmap calls.

Most activity and discussion happen in our discord but I am glad I found you here as well.

[JordanTheJet]

Also here is the PR for the docs rewrite: #5788