Actions from chat — marking, running analyses, seeing inside a post
Beyond reading, Claude can take three kinds of action through Huntit.
The same protections hold on all three. Anything that costs money runs only when you ask for it, and a whole class of these calls costs nothing at all. Marking posts and reading a tracked carousel's slides are free. Nothing reachable from chat can delete a brand, a profile or a post.
Marking posts
Discussing a post and want to keep it? Ask Claude to mark it favorite or reviewed — the same triage marks your team uses in the app, shared with everyone who can see the post. Free, instant, reversible.
Claude can read them back too, so you can ask "show me everything we starred but never briefed".
The mark is saved, not the date you set it. So "what did we mark this month?" is answered over the posts' own publish dates instead.
Running an analysis
If a post's intelligence brief, comments intelligence or custom analysis hasn't been generated yet, Claude tells you and offers to run it.
Approve, and the run behaves like clicking the button in the app: it saves to Huntit, where everyone who can see the post sees the result, and it uses your own Gemini credits at the same cents-level cost, plus your Apify account for a comments run's comment scrape.
A run starts when you ask for it. Each run's estimate lives in the app, not in the chat: Settings → Billing & costs lists all three, and a comments run also shows its estimate when you pick a depth.
On a video over the 10-minute limit, a video the AI can't read, or a text post whose text and picture weren't collected, it refuses the brief and the custom analysis outright — nothing started, nothing billed — and offers comments intelligence instead.
A Brief usually arrives in the same conversation, and waiting for it costs nothing.
Comments intelligence takes a few minutes, because the comments get collected first, so Claude starts it and checks back. A run already in progress can't be started twice.
Looking inside a post — video and carousels
Two different paths, and they cost differently.
A video
Claude can't watch video itself.
Ask Claude to "look at this video and tell me how the transitions work". Gemini watches the video and describes it: scenes, pacing, on-screen text, motion. Claude answers from that.
This uses your Gemini credits, typically a few cents.
A photo carousel
Claude can already see images, so it gets the slides — all of them, in order — and reads them itself. No AI runs on the slides, so nothing is billed for reading them.
A slide that's a video is named in its place, not watched. On an Instagram or Facebook carousel, Claude can watch one such slide when you ask: that spends one scraping credit plus your Gemini credits, typically a few cents, and Claude confirms it first.
A tracked single image or text post has nothing to watch: Claude answers from its normal detail read — the image or the caption, with its AI fields.
What's kept, and what's capped
A watch is never kept: the description lives in your conversation and nowhere else. If a watch takes long, ask Claude again later in the same chat: the finished description is held for up to a day, then deleted.
The 10-minute cap is one limit across Huntit, not a watch-only one. Over it the AI never watches the media, not even for the automatic read when the post is found.
A tracked video whose length isn't recorded, or whose file is too large for the AI, is refused the same way, before anything is billed. A carousel has no length to cap — every slide comes back.
Tracked posts and outside links
Both work on a post you already track and on a link from outside your tracking:
- A tracked post — for a carousel, the slides come with the post's normal detail read: free, instant, no AI.
- An outside link — a TikTok, Instagram or Facebook post, or a YouTube video, from a profile you don't track. Anything that might be a video is confirmed first, because a watch costs credits.
On TikTok, a photo link copied from the browser is answered straight away, free, with nothing to confirm. A share link, the short vm.tiktok.com kind, is confirmed first; if it turns out to be photos, nothing is billed.
An Instagram or Facebook link is confirmed either way. It costs one scraping credit, video or carousel, and so does a YouTube video.
If you keep coming back to the same outside profile, add it to tracking. Then everything about it is collected, analyzed, and kept.
On claude.ai the slide images are occasionally not readable in the same turn that asked for them. Ask again in the next message and they're there.
The safety rails
Huntit's guarantees don't depend on the AI: an analysis already running can't be started twice, and there is no destructive action at all. Running a finished analysis again is a new run, billed like the first.
There's no daily limit, deliberately: it's your own Gemini key, so the pace is yours to set.
In its place, Claude tells you how many paid runs you've made from chat in the last 24 hours, as the count climbs. The one hard stop is a guard against a runaway loop, set far above anything real use reaches.
The approval step you see before each action is Claude's own current behavior, on top of those.