This week's five are mostly about the parts of the work nobody wants to do twice: re-finding something you already saved, re-writing the same content for a second network, re-justifying a review nobody believes, and re-drawing a design before a line of code exists.
Design before the code exists
GlideDesign takes a short brief and returns a finished website, a mobile screen flow, or a seed pitch deck — with real written copy and generated images already in place rather than placeholder blocks. That matters for the common case where the hardest part of a first design is that it looks unfinished, so nobody can judge it.
The archive X will not give you
Twillot syncs your entire X bookmark history — past the scroll limit X imposes — into one searchable workspace, fills folders by itself, and lets you export to Obsidian, Markdown, CSV or PDF. It also keeps an offline copy, so a post survives its author deleting it. The export is the quiet feature: your saved research does not have to live wherever the tool happens to be.
A review you can trace
Signed Reviews ties each review to a real Stripe payment and signs it, then publishes a proof link anyone can open without an account. Verification aimed at the reader rather than the merchant is the only kind that actually transfers trust — and it only works for businesses paid through Stripe, which is the honest boundary of the product.
One loop for the whole content cycle
Content Drifter covers the drafting-to-posting loop in one place: on-brand drafts against your brand kit, best-time scheduling, an approval gate so nothing publishes unreviewed, and analytics that feed real engagement back into the next draft.
A launch feed you can read from code
Sunday Club publishes one hand-vetted edition a week and keeps the archive browsable — and exposes the same data over an API, an MCP server and RSS feeds, so an agent or a reader can follow it without visiting. For a directory, being machine-readable is a feature of the curation, not an afterthought.
