5 min read

Yesterday's Top Launches: 3 Tools from November 18, 2026

On November 18, three small-team products launched—a Telegram bot for cloud folders, a business assistant for UGC creators, and an AI literacy game for children—each aimed at different non-developer users.

Yesterday's Top Launches: 3 Tools from November 18, 2026

Three products went live on November 18, and the mix says something about where small teams are pointing right now. Not one of them is a new developer tools release in the usual sense. No SDKs, no CLIs, no dashboards for watching servers fall over. Instead the day brought a Telegram bot for cloud folders, a business assistant for UGC creators, and an AI literacy game for children. All three are worth knowing about, though probably not by the same person.

OctoFetch: folder links stop being a detour

Someone drops a Google Drive folder link in a chat and the small ordeal begins. Leave Telegram, open a browser, sign in, hunt for the one file that was actually needed, download it, come back, upload it again. OctoFetch (@OctoFetch_bot) exists to delete that round trip. Paste a Drive, Dropbox or iCloud folder link into the bot and it lets you browse the contents and push individual files straight into the conversation.

The appeal is obvious for anyone who lives in Telegram groups and receives folder links all day. Teams keeping project assets in Drive, freelancers handed a client’s Dropbox folder, iPhone users sharing iCloud folders with family. Same flow across all three providers, which means no relearning a different interface each time a link arrives. It is free, there are no tiers mentioned, and since it runs inside Telegram it works on mobile, desktop and web without installing anything else.

Two things the product description doesn’t answer, and they matter. How the bot gets permission to read a private folder, and what happens to your files in transit. A bot that browses your Drive is a third party sitting between you and your storage, so the access model deserves scrutiny before you hand it a link to anything sensitive. Worth trying on a low-stakes folder first.

MakeGig: the admin side of being a UGC creator

UGC creators spend a strange amount of time on work that isn’t making content. Finding brands, digging up a business email, writing a pitch, decoding an offer, remembering when a usage licence expires, chasing an unpaid invoice. MakeGig pulls those into one workspace and layers AI on top.

Brand Match surfaces companies matched to a creator’s profile, and the directory carries thousands of brand contacts across beauty, fashion, wellness and home goods, most with a published business email. Pitches are drafted from your profile plus the brand’s details, and you send them from your own Gmail or Outlook. That last detail is the right call: MakeGig never connects to your mailbox, so replies land where you expect and you keep control of the thread.

The part I find most useful is the deal analysis. Paste in an offer or an email thread and it flags what the terms actually say, what they conspicuously leave out, and the clauses worth pushing back on, like perpetual usage or exclusivity. Extracted terms are checked against the pasted text and you confirm each one. Negotiation drafts sit behind the paid tiers, and MakeGig is upfront that none of it is legal advice.

Pricing is where the honest caveats start. Free gives you 5 Brand Match swipes a month, then 3 contact unlocks, 2 AI pitches and 1 deal analysis for the lifetime of the account. That is a demo, not a plan. Creator Pro runs $1 for the first month then $6.99 monthly, Business is $14.99. The pitch that MakeGig never takes a cut of your earnings is genuinely different from how agencies work, and 228 creators listed as users is small enough that directory coverage is the thing to test before committing.

ThinkaBit: AI literacy disguised as a pet game

Teaching a seven-year-old about algorithmic bias sounds like a losing battle until you reframe it as a game about a robot. ThinkaBit gives kids aged 7 to 12 a blue robot called Sparky, who learns from their answers the way a model learns from examples. A bug named Glitch keeps slipping fakes into the training data. Professor Pixel, a retro computer in a mortarboard, explains in a sentence what just happened.

Five worlds carry the curriculum: algorithms, data and privacy, fairness and bias, cybersecurity, and deepfakes. The design is smart because the lesson is the mechanic. If you teach Sparky badly, Sparky gets things wrong, and that lands harder than any explanation of training data.

There’s a proper grown-ups area behind a PIN with progress dashboards, content controls and reports across up to five child profiles. Privacy claims are specific: COPPA and GDPR alignment, no ads, no tracking, no data selling, deletable data. The free plan covers one child profile and the first unit of every world. Family is $3.99 monthly or $24.99 yearly with a 7-day trial.

The rough edges are language coverage, where first lessons of each world are translated across eight languages and the rest fall back to English, and a subscription price that will make some parents hesitate for an app that is, at heart, a set of short lessons. The two-minute family quiz is a clever hook though, and it costs nothing to try.

Ranking data wasn’t published for any of these three at the time of writing, so the order above is the order they arrived rather than a verdict from the vote count.

Links