Core features
Stash & search
Collections that file themselves, notes for future-you, and search that works on what you remember.
Your stash is the library half of FavStash: every save, enriched and organized, retrievable by meaning.
Collections#
Collections are folders for your saves — "Recipes", "Hook ideas", "Competitor content", whatever fits how you think.
The good part: you mostly don't have to file things yourself. During enrichment, the AI reads each save and routes it to the collection that fits best, using your collections' names and descriptions as the guide. A save gets one primary collection and can belong to up to two more as secondaries — so a cooking reel with a great hook can live in both "Recipes" and "Hook ideas".
A few tips:
- Write real descriptions. The AI routes saves based on what your collection descriptions say belongs there. "Reels with hooks I want to remix" routes better than an empty description.
- You can always re-file. Manual moves win — the AI routing is a starting point, not a cage.
- Starter collections are offered on first run so you're not staring at an empty shelf.
Notes#
Every save can carry a free-text note — added from the share sheet at save time or later in the app or portal. Notes are indexed for search too, so "the one I marked steal this transition" is a working query. Write notes for future-you; future-you is forgetful.
Semantic search#
This is the headline feature. FavStash search understands meaning, so you type what you remember:
the lemon pasta thing
that reel about pricing your first digital product
@thefoodlab videos about resting meat
Search looks across everything enrichment extracted — captions, full transcripts, hooks, summaries, on-screen text — plus your own notes. It also reads intent out of your query:
- Creator handles — include
@usernameand results are filtered to that creator. - Platform words — "tiktoks about lighting" automatically filters to TikTok.
- Your phrasing doesn't have to match theirs. "Browned butter sauce" finds the video that says "beurre noisette".
Filters#
When you want to slice instead of search, filter by:
| Filter | Values |
|---|---|
| Platform | Instagram, TikTok, YouTube, Twitter/X, LinkedIn, Reddit, Threads, web |
| Collection | Any of your collections (primary or secondary membership counts) |
| Date saved | Any range — "saved this week" is a filter, not a scroll |
| Tags | Your own tag labels |
What's extracted per save#
For the curious, here's what a fully enriched save carries:
| Field | What it is |
|---|---|
| Title & metadata | Post title, creator, platform, original URL |
| Caption | The post's caption text |
| Transcript | Full spoken text, when available |
| Summary | Short AI summary of what the content is |
| Hook | The extracted opening line / stop-scrolling moment |
| On-screen text | Text that appears visually in the video |
| Visual summary | AI description of what's shown |
| Auto tags | AI-suggested topic labels |
| Your note | Whatever you wrote at save time |
| Collections | Primary + up to two secondary memberships |
All of it is indexed, all of it is searchable — in the app, in the portal, and by your AI.