collective.jbot fits in with some of this already.

[vangheem]

We're talking about hackability in sorrento. Sounds like we need to talk to you!

Here are some of the notes we made regarding where we are at with collective.jbot and how we should move forward with it::

• plone 5 compatible
• decide core, installed by default
• potentially unify interface from resource registries
• improve ace edit integration
• mockup/theme editor like
• search needs to be able to filter by types
• filter out editable items that are not useful
• template registraiton information
• icons for files
• be able to export/import of the content
• indexing the contents so searching is by contents of them, not just the name
• versioning of customizations

[ebrehault]

Actually, collective.ttw is more a discussion than an actual product.
I have initiated that after a conversion with @djay (which originally started at the Bristol Plone Conf).
I have started implementing in Diazo some parts of what we discussed (see those 2 PRs: plone/diazo#43 and plone/diazo#44 , by the way, if you see @lrowe at Sorrento, just remind him to have a look and merge them :) ).
There are probably some parts that will belong to plone.app.theming, as I think it should become our new ZMI (i.e. the place where we can easily hack Plone), so it will also cover non-design related aspects.

Regarding collective.jbot, I think it is very useful (I use it in all my Plone projects), but I an not totally happy with its approachability level (let's call it that way): it requires you know enough about Plone (what is a viewlet, where they are defined, what are the available properties and methods in the current view, etc.).
I would prefer something simpler, like for templating: no ZPT, HTML only, and for backend binding, something like a very simple Python API allowing to write very straighforward Django-like views.
But anyway, collective.jbot remains a very good tool for integrators, and the features you have listed here would be very nice.