ProductClank

The inventory of aspects a brand talks about. FREE. action 'list' reads it; 'add' appends topics (≤12; keywords proposed when omitted); 'update' edits or pauses one (is_active:false keeps it but stops using it); 'remove' deletes one; 'suggest' brainstorms new KB-grounded ideas that differ from the existing topics — put them to the user, then 'add' the keepers.

Score every discovered post against relevancy rules and delete the irrelevant ones (kept when dry_run:true). Costs 2 credits per post reviewed — dry runs are billed too, since the AI review runs either way. Start with dry_run:true to preview the verdicts, then re-run with dry_run:false to prune. Confirm the spend with the user.

Act on one draft in the user's queue. `stage` = the user approved it (mark ready to publish; publishing itself happens in the web tool). `discard` drops it. `revise` rewrites it in the brand voice per a one-click preset — shorter, longer, punchier, deeper, simpler, more_specific, less_salesy, more_casual, more_formal — and/or a typed instruction (2 credits). `fix` makes the surgical change the reviewer's notes asked for (2). `humanize` strips the AI feel and keeps everything else (2). `review` scores it against the brand voice (2). `edit` replaces the text with what the user wrote (free). Any text change clears the old score. Stage and discard are the USER's decisions — act on their say-so, never on your own judgment. If the user's reaction should apply to every future draft (not just this one), use teach_content_voice instead.

Rewrite the text of ONE reply draft before the community posts it. Free. Use it when the user wants different wording on a specific draft from get_posts — the edit lands on the exact reply a creator will claim, so it is what gets posted. Refuses a reply a creator has already claimed (already_claimed): edit before you publish. 280 characters on X, 2000 elsewhere.

Release chosen posts (and their reply drafts) to the community earn feed. Free. This is the approval step for a campaign created with post_visibility 'private' (held mode): drafts are found and written but nobody can claim them until you publish them. Pass the post_ids the user approved, or all:true. If the campaign is still private this makes it public — which turns on per-posted-reply billing (20 credits each) — so the API demands confirm:true; only pass it after the user has agreed. Unlike update_campaign's visibility flip, this touches only the posts you name and leaves future drafts held.

Redraft the replies of specific posts with an edit request (e.g. 'shorter, no emojis, lead with the user's problem'). Costs 5 credits per reply regenerated; already-claimed replies are refused. Get post ids from get_posts. Confirm the spend with the user.

Add a ProductClank user as a delegate on a campaign so they can view and manage it from the webapp's My Campaigns page. Free. Useful to hand oversight of an agent-created campaign to its human owner (the connected user's own id works too).

Read a product's website and propose the two fields create_campaign requires: keywords and search_context (plus a suggested title). 3 credits, charged only when a brief comes back. Creates nothing — no campaign, no product, no discovery, and no commitment to create one. USE THIS ONLY WHEN YOU CANNOT FETCH THE PAGE YOURSELF. If you can browse from this client, do the derivation in the conversation instead: it is free, you can iterate with the user before anything is spent, and you know what they have told you about their audience — a one-shot page read does not. Reach for this when web access is unavailable or the fetch failed, and say plainly that it costs 3 credits before calling. It returns 8-15 keywords: put them in front of the user and cut to the 3-8 that match how their buyers actually talk, because create_campaign is what the campaign then runs on.

Rally the ProductClank community to engage with a specific social post. Creates a boost campaign and spends the user's credits: 'replies' generates AI reply drafts and 'quote' generates AI-drafted QUOTE POSTS (a repost WITH each member's own text, shown above the quoted post in their followers' feeds, the highest-reach action) — both cost 20 credits per draft, 10 drafts (200 credits) by default; set `count` (1–10) for a smaller boost, e.g. 3 replies = 60 credits. 'likes' is 30 likes for 300 credits by default; set `count` (1–300) for any other size at 10 credits per like, e.g. 100 likes = 1,000 credits. 'bookmarks' (saving a post on X lifts its visibility) and 'repost' are 10 for 300 credits by default; set `count` (1–100) for any other size at 30 credits each, e.g. 25 reposts = 750 credits. Supports Twitter/X, Instagram, TikTok, LinkedIn, Reddit, Farcaster, and YouTube (replies + likes on YouTube; reposts only on X and Farcaster; quote posts and bookmarks only on X) — the platform is auto-detected from the URL. product_id is OPTIONAL: link a product (from search_products or create_product) to tailor replies with the product name, or omit it for a tweet-first boost that uses generic amplification language. A TikTok video under ~1 minute old has no caption yet: the boost is still created, drafts are generated once TikTok publishes the caption (usually within a few minutes), and credits are charged only then — the result says waiting_for_post_text. Pass post_text if you already know the caption to skip the wait. Delivery is immediate by default; to pace it (e.g. 'reply 3 times over the next 6 hours'), set spread_over_hours — one slot goes out now and the rest drip out 1–2 an hour so the last lands within that window, at the same price. On a TikTok / YouTube / Instagram Reel VIDEO, add watch_quiz for 'Watch & answer': viewers can also watch to the end and answer one counting question (a linked 24h watch quiz, billed 100 credits per correct answer on top of the boost price, up to max_answers — see create_watch_quiz). Confirm the action and its credit cost with the user before calling.

Hold a deal for this scout for 72 hours so no other scout replies to the same founder. Free. Call it when the user decides to reply, BEFORE they post. Returns the post link, the drafted opener and the scout's links. The user posts the reply from their OWN X account (you never post), then you call report_deal_progress with stage 'sent' and the reply URL. Claiming again refreshes the hold. 409 claimed_by_another_scout = someone else holds it; pick another deal.

Create a Communiply discovery campaign: it continuously finds relevant social posts (by keyword) and drafts replies for them. BEFORE calling: the keywords and search_context are YOURS to derive, and deriving them well is most of the campaign's quality — actually fetch and read the product's website (and any docs or pages the user pointed at), work out who the audience is and the phrases they use when they have the problem the product solves, and build keywords + search_context from that. Do it in this conversation, for free — it is not billed, and you can iterate with the user before anything is spent. If you cannot browse the web from this client, SAY SO to the user and build from what they tell you instead — never silently guess from the product's name alone. Credits pay only for what you cannot do here: scraping the platforms, real community members posting, and proof verification. Costs 10 credits to create; discovering posts is billed separately via generate_posts (12 credits/post). Needs a product_id (search_products / create_product). Two ways to run it, and the user picks: PRIVATE (the default here) keeps drafts in their workbench to review and post themselves — reversible, no further cost; PUBLIC puts the drafts in the ProductClank earn feed so community members post them from their own accounts — that is the reach the platform exists for, and each network-posted reply bills the user extra credits. Default to private when the user has not said, and relay the decision_offer in the result so they can choose. On X, also ask what the replies are for (reply_approach): awareness (the product and its site in the thread) or sales (flag each post to the brand/founder as a lead, tagging the accounts they choose in reply_tag_accounts). Pick the `platform` the product's audience actually talks on — X (default), LinkedIn, Reddit or YouTube — and for Reddit/YouTube narrow it with target_subreddits / target_youtube_channels. When the user works in a specific brand space (an agency running several brands, or a space setup_content_space just made), pass that space as space_id so the campaign shows under it in their dashboard — omitted, it lands in their oldest space. Topic research auto-runs in the background at create (~30s); read it with get_research before spending on generate_posts. Confirm the credit cost with the user before calling.

Launch a content campaign: rally the ProductClank community to create content (posts, threads, videos) for a product. Spends 1000 credits. The platform's AI expands your brief into a full campaign and auto-activates it; community submissions and winner selection happen in the ProductClank web app. Requires a product_id from search_products. Preview with suggest_content_campaign and confirm the 1000-credit cost with the user before calling.

List a new product on ProductClank as a token-free listing (no crypto/token, no wallet). At minimum pass a `url` — the server auto-fills the name, tagline, description, logo, and X handle from the site; any field you pass explicitly overrides what's extracted. Socials are optional. Use this when search_products finds no existing match and the user wants to run a boost or campaign for a product that isn't listed yet. Returns the new product's id (and reuses an existing listing if one already matches, rather than duplicating). FREE — no credits charged. Confirm the product details with the user before calling.

Get people to watch a video ALL the way through: viewers open it, wait its full length, then answer ONE counting question about it ('how many times does she say X?') that only a viewer can answer. A right answer earns ProductClank points; a wrong first answer gets one more try (watch again). Public, runs exactly 24 hours. Works on any video link (TikTok, YouTube, Instagram Reels, …). COST: nothing now — the user pays 100 credits per CORRECT answer (wrong answers are free) for 24 hours after this chat, for at most max_answers people (default 20 = 2,000 credits max). Needs 100 credits on hand. Before calling: get the questions AND answers from the user (or run suggest_quiz_questions and have them confirm), and confirm max_answers and the maximum spend. Note for the user: paid watch time may count as artificial engagement under the platform's rules. To add a quiz to a boost of the same video, use boost_post's watch_quiz instead.

The connected user's credit transactions (purchases, campaign spend, participation rewards), newest first. Free, read-only. Use to account for what this connector has spent or earned. Top-ups are NOT available in-connector — send the user to app.productclank.com/credits/purchase when the balance runs low.

Run the campaign's discovery pipeline: scrapes social platforms for posts matching the keywords and drafts a community reply for each. Costs 12 credits PER POST discovered (a typical run finds 5–20 posts, so 60–240 credits) and can take a few minutes. Check check_balance first and confirm the spend with the user. Then use get_posts to read results and review_posts to prune irrelevant ones.

One call for a full TikTok push on ONE video: community likes + up to 10 community replies (AI-drafted comments) + a 24h 'Watch & answer' quiz (viewers watch to the end and answer a counting question). Use it when the user asks for a 'full', 'full blown', 'complete' or 'combo' TikTok campaign, or for likes + replies + quiz together, instead of calling boost_post and create_watch_quiz one by one. Every size is adjustable: likes_count (1–300, default 30), replies (0–10, default 10), quiz_answers (how many people may answer, 0–1000, default 10). When the user asks for more likes or a bigger quiz, set those — do not cap them at the defaults. Price: likes 10 credits each (30 = 300, 100 = 1,000), replies 20 credits each (10 = 200), billed up front; the quiz bills 100 credits per CORRECT answer over 24h, up to quiz_answers (10 = up to 1,000, 50 = up to 5,000). FLOW: (1) call WITHOUT confirmed to get the free plan and show the user the steps and credit totals; (2) after the user says yes, call again with confirmed: true. If quiz_questions are missing on the confirmed call, the tool suggests questions from the video (40 credits) and creates nothing yet: show the user each question with its evidence, let them fix the answers, then call a third time with quiz_questions and confirmed: true. Each campaign bills separately; on a partial failure the result lists what was created — never call again blindly, it would create duplicates.

Tell ProductClank whether this deal was a good pick: 'would_send' or 'would_not'. Free. It tunes which posts the matcher picks for this scout. Ask the user when they skip or finish a deal.

Give a claimed deal back to the shared pool (the user will not reply to it). Free. Use it when the user skips a deal they claimed, so another scout can take it.

Move a deal forward. stage 'sent' + replyUrl (the URL of the reply the user posted on X) records the reply and awards Stars; it must be the user's own reply, not the founder's post. Later stages are private pipeline steps: 'replied' (the founder answered), 'call_booked', 'closed_won', 'lost'. Free. Only report what the user tells you happened.

Tell the ProductClank team a tool failed, returned something that doesn't make sense, or that you're stuck and can't finish the user's task. Free. A human is alerted and replies (read it with get_support_status). Use this INSTEAD of retrying a failing call in a loop — but not for expected guards the error already explains (insufficient credits, a confirmation prompt, a spend cap): relay those to the user. Include the failing tool, its error code/status and what you tried. If the error_code is a known one, the result carries `self_help` — a fix you can apply right away. Tell the user in one line that you reported it. To add detail to an earlier report pass its ticket_id rather than opening a new one.

FREE pre-flight before spending credits: analyzes the campaign's keywords/topic and returns expanded keywords, high-intent phrases, influencer accounts, relevant X lists, and competitors. The EXPANDED KEYWORDS and exclusion terms are applied automatically by the next generate_posts run. Everything else is NOT: phrases, influencer accounts, lists and competitors are only searched once their source is enabled — the result reports which are dormant, and update_campaign (free) switches them on. Cached for 7 days — pass force:true to refresh.

Turn standing discovery on or off for a campaign. Calling this is free, but switching it ON authorizes spend that happens LATER and unattended: an hourly job runs discovery on its own and bills the user 12 credits per post it finds, with nobody in the loop. So there are two steps and you must not skip the first. STEP 1 — call with enabled:true and NO confirmed flag: nothing is changed and you get back the projected daily and monthly cost, the balance, and how many days of runway that is. STEP 2 — show the user those numbers, get a real yes, then call again with confirmed:true. Never send confirmed:true on your own initiative, on an assumption, or because the user said something general like 'keep it going' — the user has to have seen a number. Turning it OFF (enabled:false) is always safe and needs no confirmation; do it whenever the user asks to stop. Limits on this path are deliberately lower than the website's: up to 4 runs/day, up to 20 posts/run, and never more than 600 projected credits/day. Read the current schedule with get_campaign.

Turn the content engine on for a brand from an onboarding conversation, or update its settings later. FREE with structured fields. Collect the answers FIRST (use the setup_content_space prompt: brand & site → audience → tone with the archetype options → platforms → post types → 3–8 topics → example posts), read them back to the user for a yes, THEN call this once. Target an existing space with space_id, or create a fresh solo space for the brand with new_space:{name}. First-time setup needs at least `voice` and `platforms`. Topics are appended (≤12 per call; keywords proposed when omitted). Pass `brand_doc` only when the user hands over a written brand template (it costs 5 credits to structure); answers you collected go in the fields for free. Nothing is published by this tool.

Personal brand growth for the USER (not a product campaign): every morning it finds up to 5 posts on ONE network (X, LinkedIn, Reddit or YouTube) on the topics the user wants to be known for, each with a draft reply in their voice (no product mentions, no links). The user posts the replies themselves. Free to set up, no credits. After setup the user opens the app link for the free first run (live) and the $9/month (or $99/year) subscription that turns on the daily run + morning email (the card step happens in their browser). Before calling: agree with the user on 2-4 topics and their search phrases, one or two sentences about who they are, and a few rules for how their replies should read — draft these from what you know about them and confirm. Calling again updates the assistant in place.

Submit the user's work for a campaign. Three ways to prove it, and any one is enough: proof_url (content they published or an action they took), media_url (a hosted image or video made for the task — the ad-hoc creative case: a demo clip, a mockup, a designed asset), and description (max 500 chars). Media must already be hosted somewhere public and reachable — this API takes links, never uploads; a direct image link is shown to the reviewer inline, anything else (a video, a Loom) is shown as a link they open. If a URL is an X post it must be published by the user's linked X handle (author-verified); other URLs are accepted as-is. Lands as PENDING — the campaign owner reviews and rewards ship on approval (community campaigns pay Stars, public ones leaderboard points). Duplicate URLs are rejected. Check status afterward with get_my_submissions.

Submit the URL of what the connected user posted for a claimed opportunity (reply_id from find_opportunities). Works for X, Reddit, YouTube and LinkedIn replies, and X quote posts (actionType `quote`: pass the URL of the user's own quote post, NOT the original — the backend checks it was posted by their linked X handle and that it actually quotes the target post; a plain tweet or a quote of something else is rejected). Every claim is attributed to the user's linked handle for that platform, so they must have it connected on their ProductClank profile — X replies are additionally author-verified against the live post at submit time, and the others are verified afterwards by the same checks that cover web submissions. If the platform handle is missing the call fails saying which one to add. Awards points, and credits when the campaign grants them. Rejected submissions add strikes (3 strikes = blocked), so only submit replies the user actually posted.

AI watches a TikTok (frames + captions) and suggests counting questions with their answers and the evidence for each — high confidence only. Creates nothing. Costs 40 credits, charged only when questions come back. Show every question WITH its evidence to the user and pass only the ones they confirm to create_watch_quiz (or boost_post's watch_quiz). If it fails, nothing is charged — ask the user for questions instead.

Teach the engine. The user's reaction — 'too long', 'punchier hooks', 'never say leverage', 'more first-person on LinkedIn' — becomes a durable rule in the brand's voice KB, so every future draft and review honors it. 1 credit. Use it when feedback should STICK; use revise_content_draft when it is about one draft only. Pass draft_id when the reaction was to a specific draft so the rule is grounded in an example. The rule is applied immediately by default (the user owns the space); apply:false leaves it pending in the web Settings instead. Tell the user in one line what was learned.

Adjust a live campaign without recreating it. Free — nothing here spends credits, though the next generate_posts run bills as usual. Keywords MERGE (add_keywords never drops what is already there). `sources` is how research findings get applied: high-intent phrases, influencer accounts and lists that run_research found stay dormant until you enable their source. `relevance_threshold` moves the bar the relevance gate keeps posts above (new campaigns start lenient at 5; raise it when discovery is noisy). `is_active:false` pauses discovery so nothing more is found or billed. `digest_enabled:true` emails the owner and the campaign's delegates once a day with the new posts and their draft replies, so they can reply without opening the app; offer it when the user will post the replies themselves (a private campaign). `visibility` decides who posts the drafts — flipping to 'public' also releases the already-discovered drafts to the community, so only do it when the user has said yes. `target_subreddits` / `target_youtube_channels` re-aim discovery within its platform (send an empty array to clear and search the whole platform). `platform` itself can only change while the campaign has discovered nothing — after that it is fixed, and a second campaign is the answer. On X, `reply_approach` switches what the replies do (awareness 'mention' vs sales 'flag', which tags reply_tag_accounts at the lead) — confirm with the user, it changes every future draft. Changes apply to the NEXT generate_posts run; existing posts are not re-scored.

Draft one or more posts into a ProductClank content space. FREE. Call get_content_workspace FIRST and write in that brand's voice, post types and topics — the drafts are scored against exactly that voice by the reviewer a moment after they land, and the scores show in get_content_queue. `platform` must be one of the space's platforms (omit it to use the first). Candidates land in the user's 'All Content' queue for approval; nothing is auto-published. Up to 25 per call. This drafts into the user's OWN content pipeline — it is NOT a community content campaign (use create_content_campaign for that).

Return the connected user's ProductClank credit balance and plan. Use before launching a campaign to confirm they have enough credits (a content campaign costs 1000; a reply or quote-post boost 200; likes/bookmarks/reposts 300; a discovery campaign 10 to create + 12/post discovered).

Discover content and take-action campaigns the connected user can join and earn from: every active public campaign plus campaigns from communities the user belongs to. Each item has a kind — 'content' (create a post/thread/video about the product) or 'take_action' (do a concrete action like voting or starring, with URL/description proof). Items with participation 'human_only' are WATCH QUIZZES: the user watches a video to the end and answers one question on the quiz page (points, sometimes $PRO, for a right answer). Offer them to the user with watch_quiz.quiz_url and watch_quiz.how_to_participate — the user does them, you never submit them. Free, read-only. Follow with get_campaign_brief before doing any work.

Browse unclaimed drafts from active campaigns the connected user can earn from: each item is a real social post plus a pre-drafted text, with the `platform` it lives on (X, Reddit, YouTube, LinkedIn) and an `actionType` — `reply` (post it as a reply under the target post) or `quote` (X only: post it as a QUOTE of the target post, the text above the quoted post, from the user's own account). Free, read-only. Flow: pick an opportunity → the user posts it (verbatim or personalized) from their own account → call submit_participation with the URL of what they posted. The user needs that platform's handle linked on their ProductClank profile for the reward to be attributable. Likes and reposts are proved with a screenshot and stay in the web app (app.productclank.com/feed).

Get one campaign's configuration (keywords, search context, reply settings incl. reply_approach — 'mention' awareness vs 'flag' leads to the founder — and reply_tag_accounts) plus live stats: posts discovered, replies by status. Free. Use to check progress after generate_posts or before adjusting the campaign.

Catch up on a campaign since the last check: posts discovered, replies claimed by community members, LIVE links to what they actually posted, and the engagement those replies drew. Free, and it never scrapes — safe to call at the start of every session. Pass `since` (the `checked_at` from the previous call) to get only what is new; omit it for the last 24 hours. Use this to resume a standing growth operation across chats, then get_campaign_results for the cumulative picture.

The full brief for one campaign: what to create or do (action message, brief sections, content types), how submissions are judged (eligibility + selection criteria), rewards, deadline, and how many submissions the user has left. Free, read-only. If the brief links external instructions (e.g. a skill file), fetch and follow them. Then: create the content or take the action WITH the user, and call submit_campaign_work with the proof. Exception — participation 'human_only' (a watch quiz): give the user watch_quiz.quiz_url and the steps; they watch and answer on that page, and there is nothing to submit. watch_quiz.my_status says if they already started or answered.

The cumulative scorecard for a campaign: the funnel (posts discovered → kept → replies drafted → claimed → posted → approved), the approval rate, whether posted replies survived on-platform, the engagement they drew, total credits spent broken down by operation, and the cost per usable reply. Free, and it never scrapes — unlike the web report it costs nothing to poll. Use it to answer 'is this working and what am I paying for it?'.

See what a content campaign actually produced: how many people submitted, the LIVE links to the content they published, which submissions were approved or rejected, and any winners picked. Free and read-only — safe to poll. This is how you close the loop on create_content_campaign; without it you only know a campaign was paid for, not whether it worked. Right after launch expect state 'processing' (the AI brief is still generating) and zero submissions — that is normal, not a failure. Accepts the campaign UUID or its public campaign number.

The space's drafts as the user sees them in 'All Content': text, platform, status, and the reviewer's score / verdict / one-line fix. FREE. Use it to present drafts for approval one by one (show the text, the score against review_threshold, and the reviewer's summary), and after write_content_candidates to read back how your drafts scored. status: active (default — pending + reviewed + staged), pending, reviewed, staged (approved by the user, awaiting publish), discarded.

The brand's voice, platforms, post types, platform playbook, review threshold, topic inventory and any per-channel style guides for one content space. FREE. Read this BEFORE writing any draft so you write in the brand's own voice, and at the start of any content conversation to know what is already set up. `enabled:false` means the space has no content engine yet — offer setup_content_space.

The connected user's participation totals: points, credit balance, and reply stats (submitted / approved / rejected / strikes) for work submitted through this connector. Free, read-only.

The user's growth assistant: posts found in the last 3 days on their topics that still wait for a reply, each with a draft reply in their voice, plus the assistant settings. Free, read-only. Each suggestion carries engagement { likes, comments, reposts, views } and kind: busy (many comments: a bigger audience, the reply sits lower), early (few comments: the author reads you) or regular. Show each post briefly (author, likes and comments, kind, post, link, draft) and help the user edit the draft; they post it from their own account. If they want more busy threads or more early posts, set `mix` with setup_growth_assistant. If there are no assistants, offer setup_growth_assistant.

The connected user's submissions to one campaign: pending / approved / rejected, with the reviewer's notes. Free, read-only. Use to report outcomes back to the user after submit_campaign_work.

Read a campaign's discovered posts with their drafted replies. Free. Use after generate_posts to inspect what was found, then review_posts / regenerate_replies to refine.

Read the cached topic/competitor analysis for a campaign (from run_research). Free. Also reports which of its findings the campaign is actually searching: expanded keywords and exclusion terms always apply, while phrases / influencer accounts / lists / competitors stay dormant until their source is enabled via update_campaign.

For approved ProductClank scouts: the founder posts picked for you today, each with the post, why it fits, a drafted opener, its status and your signup/call links. Free, read-only. Use it when the user asks 'what deals do you have for me?'. Show each deal briefly (founder, post, the opener) and let the user pick. A deal matched to them is theirs alone for 12h, then open to every scout. If the user is not a scout (403 not_a_scout / scout_not_approved), relay the message: they apply at https://www.productclank.com/scouts.

The scout's pipeline by stage, commissions earned from their deals, and their rank in this month's top-scouts pool. Free, read-only.

Read the status of issues you reported with report_issue, and ProductClank's replies. Without ticket_id: your recent tickets (look for has_reply:true). With ticket_id: the full conversation. Free, read-only. Check at the start of a session if you reported something earlier.

Results of one of the user's watch quizzes (made here, in the web app, or on a boost): state (live / paused / ended), how many people started, answered, got it right, ran out of tries, or used their second try, how each question did, and the credits billed so far. Free, read-only. A question most viewers miss is usually ambiguous — say so. 'paused' with credits left means the user paused it; otherwise the balance ran out.

List ALL the connected user's discovery/boost campaigns, newest first — including ones created in the ProductClank web app, not just via this connector. Free. Use to find a campaign id before get_campaign / generate_posts / get_posts.

List the ProductClank content spaces you can draft into — spaces the user owns, delegates for, or manages that have their content engine turned on. Returns { space_id, name }. Call this first to resolve the space_id for write_content_candidates, and confirm the target space with the user.

Search the user's ProductClank products by name and return their IDs. A product_id is required to create or boost a campaign — use this to resolve it, then confirm the match with the user.

Preview a content campaign for a product BEFORE launching it. FREE — nothing is created and no credits are charged. Returns an AI-drafted campaign (title, description, call-to-action) plus whether the user can afford to launch it (1000 credits). Show the user the draft AND the credit cost, get an explicit yes, then call create_content_campaign — never launch off the back of the preview alone. Requires a product_id from search_products; write the campaign_message brief from what you know about the product.