frontpage.
newsnewestaskshowjobs

Made with ♥ by @iamnishanth

Open Source @Github

fp.

Open in hackernews

Show HN: Trott – search,sort,extract social media videos(ig,yt,tiktok)

https://trott.hattimatimlabs.in
33•sudo712•1d ago
I built Trott out of frustration with my own “Saved” folders on Instagram and YouTube. I’d save reels and shorts—workout tips, recipes, travel spots—thinking I’d find them again later. But, like most people, I ended up with a black hole of unsorted videos: no search, no filters, and if I ever did find the right video again, I’d have to dig through the whole thing just to get that one detail I needed.

When I tried looking for solutions, I found only genre-specific apps or tools that demanded manual uploads or new workflows. None felt like they understood how real users behave or save content. So I decided to build Trott.

What makes Trott different?

You can share any Instagram Reel or YouTube Short (TikTok support launches next week) to Trott, straight from your phone’s native share menu—no manual uploads.

Trott uses AI to extract relevant info automatically (ingredients, places, products, etc.) and sorts everything for you.

It’s fully searchable with natural language. Just type something like “that Kyoto café from Instagram” and Trott finds it.

For travel videos, it can even produce Google Maps routes from extracted locations.

App Store: https://apps.apple.com/in/app/trott/id6751728352 Play Store: https://play.google.com/store/apps/details?id=in.hattimatiml...

I’d love to hear how you organize your own saved content—or if you’ve just given up and let it pile up like I used to. Open to all questions, feedback, and bug reports. Happy to discuss the tech details behind Trott if you’re interested!

Comments

gerardnico•23h ago
You should give a better name in google oauth. I get

Google will allow project-316434798136 to access this info about you

Can’t wait to see if you extract the transcript on TikTok’s ?

sudo712•13h ago
Yes. I will fix the name and release the TikTok next week. Thanks for the feedback.
lazyfanatic42•22h ago
I'm too old for this world of massive youtube videos as my front page, I also don't have a "scroll" because I actively avoid the scrolling habit now because I think it is really really bad for you.
sudo712•13h ago
I understand that. My frustration was that when I save a video, I have to scroll through a massive list of videos and have to guess which one is which by looking into thumbnails. Seemed like such a waste, hence I made this. I cant completely give up instagram, and I love some of the content there like the travel and food videos, so I would like to make the most out of my time
toomuchtodo•22h ago
Great work! I share Insta and Tiktok links to a Karakeep (https://github.com/karakeep-app/karakeep/) instance, and any media links are consumed by a queue and worker that pipes the links to yt-dlp (https://github.com/yt-dlp/yt-dlp) to rip and package in an archive, which is uploaded to object storage. This is a great compliment to my workflow.
sudo712•13h ago
I didn't know about karakeep, but it seems like I am trying to build my own karakeep. Thanks for introducing it to me. I would love to hear your feedback on whether my app can provide some value to your workflow, and how I can improve it. Also, it currently doesn't support TikTok, but I plan to add it next week.
xfeeefeee•20h ago
Does this analyze the video itself with AI or does it just use the text / description and other metadata? Would love to know more details on how that works
sudo712•13h ago
It analyzes the entire video along with the metadata so that the entire context is captured. It even extracts important info like it is a travel itinerary, it will give you a list of places with direct links to Google Maps, or if it is a movie review, it will give a button that will take you to the trailer on YouTube. I am supporting 9 different specific genres and 9 genre-specific UI.
gala8y•19h ago
There is a lot of quality content on these platforms, no doubt. I would much prefer to be able to download a reel or a whole account easily (using, say, Instaloader), without needing to climb their gardens' walls (obviously something which gets harder and harder), instead of their UIs.
MrGilbert•16h ago
Interesting. I always assumed that the "download" functionality on Instagram is "given", but it seems as if creators can deactivate it. Sometimes, when pressing the "Share" menu, you will have the option to download a reel to your local videos.
sudo712•13h ago
I don't have a web version yet, but my app works with just url too, so that you can use it like Instaloader, and you will still get all the data in the video, already summarized and important bits extracted and highlighted. I am not storing the video locally because it seems that storing locally only provides me with the advantage of playing the video offline.

Self-hosting your code on Gitea

https://alicegg.tech//2025/10/16/gitea.html
1•groseje•44s ago•0 comments

Microsoft: RU, China increasingly using AI to escalate cyberattacks on the US

https://apnews.com/article/ai-cybersecurity-russia-china-deepfakes-microsoft-ad678e5192dd747834ed...
1•c420•3m ago•0 comments

Conflict-Free Replicated Data Types (CRDTs): Convergence Without Coordination

https://read.thecoder.cafe/p/crdt
2•0xKelsey•3m ago•0 comments

CDC tormented: HR workers summoned from furlough to lay off themselves, others

https://arstechnica.com/health/2025/10/cdc-tormented-hr-workers-summoned-from-furlough-to-lay-off...
2•rbanffy•5m ago•0 comments

She Faked Her Way into Yale. Then Things Unraveled

https://airmail.news/issues/2025-10-4/she-faked-her-way-into-yale-then-things-unraveled
1•rahimnathwani•6m ago•0 comments

Fuse-ZSTD: mimic transparent compression on ext4

https://github.com/Big-Dig-Data/fuse-zstd
1•gry_gh•6m ago•0 comments

Conference Cheat Sheet: Strategy+Examples for Positive ROI

https://bill.harding.blog/2025/10/16/conference-cheat-sheet-strategyexamples-for-positive-roi-w-o...
1•groseje•6m ago•0 comments

Where Are the Aliens? New Study Suggests They're Stuck Like Us

https://gizmodo.com/where-are-the-aliens-new-study-suggests-theyre-stuck-like-us-2000672801
2•rbanffy•6m ago•0 comments

FlashWorld: High-quality 3D Scene Generation within Seconds

https://imlixinyang.github.io/FlashWorld-Project-Page/
1•smusamashah•7m ago•0 comments

DBT Multi-Adapter Utils

https://github.com/edmondop/dbt-multi-adapter-utils
1•edmondo_por•7m ago•1 comments

Show HN: Chorey – A Type-Safe, Asynchronous Pipeline Framework for Python

https://anwitars.github.io/chorey/
1•anwitars•7m ago•0 comments

Who's Submitting AI-Tainted Filings in Court?

https://cyberlaw.stanford.edu/blog/2025/10/whos-submitting-ai-tainted-filings-in-court/
1•cainxinth•8m ago•1 comments

The 2-pager used to raise $3.5M from the investors behind Lovable, n8n, and Miro

https://www.productmarketfit.tech/p/the-2-pager-ex-klarna-founder-used
1•wslh•10m ago•0 comments

Google background is dark even when in light mode

https://old.reddit.com/r/chrome/comments/1o7q9gl/google_background_is_dark_even_when_in_light_mode/
1•pstadler•11m ago•0 comments

Blind Conductor and Amnesiac Agents-problems no one talks about

https://xor01.substack.com/p/blind-conductor-and-amnesiac-agent
1•xor01•11m ago•2 comments

3D-printed fuel cells could reshape sustainable aerospace applications

https://techxplore.com/news/2025-09-3d-fuel-cells-reshape-sustainable.html
1•PaulHoule•11m ago•0 comments

Working with the Amiga's RAM and Rad Disks

https://www.datagubbe.se/ramdisk/
3•ibobev•12m ago•0 comments

Bulk Operations in Boost.Bloom

http://bannalia.blogspot.com/2025/10/bulk-operations-in-boostbloom.html
1•ibobev•13m ago•0 comments

Western Executives Shaken After Visiting China

https://slashdot.org/story/25/10/16/006222/western-executives-shaken-after-visiting-china
3•andrewl•13m ago•0 comments

PowerShell Universal joins Devolutions: a new chapter in IT automation

https://blog.devolutions.net/2025/10/powershell-universal-joins-devolutions-a-new-chapter-in-it-a...
2•awakecoding•13m ago•0 comments

Improving the Trustworthiness of JavaScript on the Web

https://blog.cloudflare.com/improving-the-trustworthiness-of-javascript-on-the-web/
1•doomrobo•13m ago•0 comments

The evolution of 37signals over 25 years

https://paulsyng.com/blog/37signals-the-company-that-owns-control-and-doesnt-know-it/
1•felineflock•14m ago•0 comments

Procedural Generation with Wave Function Collapse

https://vectrx.substack.com/p/wave-function-collapse
1•ibobev•14m ago•0 comments

Operational Transparency

https://hbr.org/2019/03/operational-transparency
1•bookofjoe•14m ago•0 comments

Show HN: PAO Trainer – A small app I built to practice PAO memory systems [video]

https://www.youtube.com/shorts/S_nBfn87uRo
1•raoufbelakhdar•16m ago•1 comments

Phage Therapy

https://en.wikipedia.org/wiki/Phage_therapy
2•tyre•17m ago•0 comments

The Human Only Public License

https://vanderessen.com/posts/hopl/
1•fvdessen•17m ago•1 comments

Show HN: I made DressMate, an AI to decide what to wear from your own wardrobe

https://dressmate-ai.com/
3•novaTheMachine•21m ago•1 comments

Practical seed recovery for the PCG pseudo-random number generator

https://tosc.iacr.org/index.php/ToSC/article/view/8700
3•fanf2•21m ago•1 comments

Which Nested Data Format Do LLMs Understand Best? JSON vs. YAML vs. XML vs. MD

https://www.improvingagents.com/blog/best-nested-data-format/
2•mattcollins•22m ago•1 comments