huntitHelp center

Playbooks — the multi-step power moves

Single questions are the cookbook — that's Usage ideas. Playbooks are the level above: one ask that ends in a finished deliverable, with three or four pulls behind it.

Read each one this way: the line in quotes is what you type. Everything after it is what Claude goes and fetches on its own to answer — not steps for you to run.

The power combinations

Winner → script pipeline

"Write me 3 scripts that reuse this winner's recipe, in our client's voice, using the audience's own words."

Claude searches the winners, opens the strongest one for its transcript and hook, reads its Brief for the reusable moves, and pulls the audience's own phrasing out of its comments.

The one-page competitive audit

"Build me a one-page audit: who's winning, on which formats, with what hooks — and where's the whitespace?"

Claude maps the client's tracked competitors, pulls each one's winners, and reads each profile summary.

A content calendar built from data

"A 4-week calendar for [client], one hook and one angle per post, anchored in what wins and what the audience asked for."

Claude takes the last 90 days of winners grouped by content group, and crosses them with the audience's desires and questions.

The complete client update

"The monthly email, 3 talking points for the call, and a 'what we test next' line."

Claude reads the stored report, the month's winners, and today's radar.

Trend sprint → generation

"Break down the fastest riser on the radar right now, then give me a storyboard I can hand to my video tool."

Claude takes the radar's fastest riser, opens its detail, watches the video, and reads the Brief's reusable moves. One conversation, from signal to asset.

The audience backlog

"20 content ideas, each mapped to the audience layer it answers."

Claude reads comments intelligence across a profile's winners and pools the objections, desires and questions.

The hook bank

"A swipe file of 30 hooks, grouped by pattern, each with its source post."

Claude searches the winners and extracts every Hook and Visual hook.

Format diagnosis

"Why aren't we winning on [format], who is — and what do they do differently?"

Claude compares your win rate and content mix against each competitor's.

The repurpose engine

"Give me this winner as a TikTok script, an Instagram carousel outline, a YouTube Shorts hook, and a text post — the same insight, native in each place."

Claude reads the one post and works from it.

Video mechanics breakdown

"Watch [link] and break down the editing — cuts, pacing, on-screen text, transitions. Then apply the same moves to our concept."

Claude has the video watched and reasons over what came back.

The carousel teardown

"Read this carousel's slides and map how it builds — the promise on slide one, the turn in the middle, the ask at the end. Then outline ours on the same skeleton."

Claude reads the slides itself. It's the cheapest playbook here: no AI watch runs on a carousel, and on TikTok the read is free.

Across all your clients

Your whole portfolio is searchable at once — three moves agencies with several clients get for free:

  • "Across ALL my clients, which content group is winning the most this month?"
  • "What reusable move keeps appearing in winners across different clients?" — a transferable playbook.
  • "Show me a winner from client A's market that would translate to client B." — cross-pollination.

A weekly rhythm

  • Monday: "what won last week across my clients — and what's rising this morning?"
  • Report day: "summarize the new report for [client] and draft the email."
  • Before filming: "brief the person shooting: what's winning in this market + the hooks worth trying."
  • Before a client call: "brief me on [competitor] + our own numbers."

The habits that get better answers

  1. Name things the way Huntit does — clients, brands, profiles — so Claude finds them on the first try.
  2. Ask for the source — "which posts is this based on?" makes Claude show the winners behind its claims.
  3. Read before you generate — pull the recipe (Brief, transcript, comments) BEFORE asking for assets.
  4. Go wide, then narrow — start with search or the radar, then deepen one thread: detail → Brief → comments → draft.
  5. Approve costs consciously — analyses and video watches use your agency's own Gemini credits; a carousel's slides run no AI at all. Claude tells you an action is billable before it runs; the per-run estimate lives in the app, not in the chat.

Keep it honest

Huntit shows what worked — it never guarantees what will. Ask for observations first and generation second, and the answers stay grounded: "6 of 14 winners were Storytelling" is a fact you can build on. Costs and saving rules live in Actions from chat.