compare_competitors_timeline
ChatGPTSide-by-side monthly metrics for multiple competitors over a rolling window. Summary metrics only (postCount, avgEngagement, medianEngagement, follower boundaries). Caps: 6 competitors, 12 months.
get_actionable_summary
ChatGPTFetch the user's weekly actionable summaries (Mon–Sun recap of competitive activity). Newest first, cap 5.
get_ads_in_range
ChatGPTIndividual ad creatives in a date range (parallel to get_posts_in_range). Filter by ad START date — same semantics as the monthly digest. Returns full creative copy (bodyText, title, CTA, dates, publisher mix / country mix, EU reach, preview URL). Cap 50, newest first. Ad tracking needs the Basic plan or higher.
get_alerts
ChatGPTFetch the user's real-time competitor activity alerts with filters. Caps at 50 alerts.
get_battlecard
ChatGPTLatest generated battlecard for one competitor — the synthesized strategic playbook (tagline, positioning, threat + rationale, competitive frame, audience overlap, messaging skeleton, win patterns, channel war room, anti-patterns, proof points, when-not-to-fight, top action, scenarios, commitments, external signals). Includes a compact 12-month trajectory of older cards (max 5) so consumers can see how the read has shifted. Battlecards needs the Starter plan or higher.
get_competitor_ads
ChatGPTPaid-advertising data for one competitor across Facebook/Meta Ads Library and Google Ads Transparency Center. Per channel: tracked URL, 12-month plan-window aggregate (totalAds, activeAds, EU reach for Facebook, country breakdown for Google, format/CTA mix, daily history of last ~30 days), and recent top creative samples (max 15) with bodyText / title / CTA / dates from the last 3 monthly digests. Channels the competitor isn't tracked on return null. Ad tracking needs the Basic plan or higher.
get_competitor_ads_timeline
ChatGPTPer-month paid-ad timeline (parallel to get_competitor_timeline for posts). Per channel + month: adCount, newAdCount, activeAdCount, EU reach (Facebook) / country breakdown (Google), format & CTA mix, cadence by day of week, top keywords. includeSamples=true adds up to 5 top creative samples per month (default false). Cap 12 months. Ad tracking needs the Basic plan or higher.
get_competitor_details
ChatGPTFetch the full CompetitorSummary (all platform metrics + followersHistory) and the 50 most recent posts for one competitor, plus links (the tracked website / social / Trustpilot / ad-library URLs, null = not configured). Returns null if the id is invalid or not owned by the user.
get_competitor_profile
ChatGPTReturn the latest AI-generated qualitative profile (company context, industry context, messaging pillars, tone of voice, market dynamics). Use for narrative questions; for numeric metrics use get_competitor_details.
get_competitor_reviews
ChatGPTConsumer-review data for one competitor (Trustpilot today). Returns: tracked URL, 12-month plan-window aggregate (totalReviews, trustScore, avgRating, rating distribution, daily review-count history), and recent sample reviews (max 15) with text / rating / language / companyReply (the business's own response to the review, null when none) from the last 3 monthly digests. Channels not configured on the competitor return null. Review tracking needs the Basic plan or higher.
get_competitor_reviews_timeline
ChatGPTPer-month review timeline (Trustpilot). Per channel + month: reviewCount, avgRating, trustScore, ratingDistribution, cadence by day of week, top keywords. includeSamples=true adds up to 5 sample reviews per month (default false). Cap 12 months. Review tracking needs the Basic plan or higher.
get_competitor_suggestions
ChatGPTList the user's pending competitor RECOMMENDATIONS — companies the system suggests they track, each with a rationale, score, and discovered profile/ad/review URLs. Read-only; does not add anything or run discovery. Returns up to 20, highest-scored first.
get_competitor_timeline
ChatGPTPer-month strategic timeline for one competitor (post counts, engagement, top 10 posts, posting cadence, media-type breakdown, top keywords/hashtags). Newest month first. Capped at 12 months.
get_content_policy
ChatGPTFetch the user's Studio content policy — the per-brand tuning the content-calendar planner obeys: per-channel posting cadence (postsPerWeek) and posting windows, the format-mix weights (the single source of truth for how much of a week is promo, social proof or blog traffic), and the experiment budget (caps.exploreShare). Each tunable is returned as { value, source, aiRecommended } so you can see the user's setting next to the AI recommendation, plus each mix row's capability and locked/lockedReason (a format that cannot be scheduled at all right now). Nothing caps a weight: the mix is the brand's own decision, and the planner schedules what the source material can cover. caps.promoDensity / caps.reviewsPerWeekTotal are derived from the mix and read-only. Read-only; use the Studio app to edit or regenerate it. The Studio needs the Starter plan or higher.
get_insights
ChatGPTFetch the user's AI-generated insight cards across all competitors with filters. Caps at 100 cards.
get_learning_report
ChatGPTThe decision-learning report for the brand on this connection: the share of AI drafts sent UNEDITED over the last 30 days (with a 95% Wilson interval) and a 12-week series, the weekly median edit distance, the backtest metrics of the promoted learned models (pairwise accuracy, ECE, trained date; null when none), learned preferences with counts (approval by hook type / tone / CTA / topic, only values with 5+ decided drafts), learning warnings shown / accepted / overridden / held out with the measured outcome lift of each group, and per-area data sufficiency. Every number carries its counts; null means NOT MEASURED, never zero. Read-only. Studio plans (Starter and up).
get_own_ads_performance
ChatGPTDistilled performance of the user's OWN connected ad accounts (NOT competitor ads): totals (spend, impressions, clicks, CTR, CPC, CPM, conversions, cost/conversion, ROAS), a per-platform breakdown, top active campaigns, and week-over-week deltas. Default window is the trailing 30 days; pass fromDate/toDate (YYYY-MM-DD) to change it. Returns { hasAds:false } when no ad account is connected. Use this to answer "how are my ads doing".
get_playbook
ChatGPTFetch the user's OWN strategic Playbook — the living strategy document about their company (not a per-competitor artifact). Returns full prose for every written section across the foundation / strategic / tactical / performance layers (positioning, ICP, brand, channel strategy, KPIs, plays, etc.) with live metric figures already resolved into the text. Also reports which sections are stale (inputs changed since last write), which are empty, and which have a pending AI suggestion. Use this to load the user's full strategy into context before advising on positioning, messaging, or planning. Optionally filter to specific sectionIds. Basic plan or higher required.
get_posts_in_range
ChatGPTReturn posts for a competitor filtered by date range, platform, and media type. Dates must be ISO-8601 (YYYY-MM-DD or full timestamp). Caps at 100 posts. Omit platform and mediaType and one call returns EVERY channel and media type in the window, each post carrying its own platform — that is the default way to ask; set them only for a deliberate slice, and never loop the same window across platforms or media types. An empty result with a filter set means "none of THAT kind", not "no posts": when your filter matches nothing the tool re-runs the same window WITHOUT filters and returns those posts, marked filterMiss.recovered=true (platform/mediaType then echo back as 'all' because that is what posts[] now covers) — use it as-is, do not re-call the window, and never report the period as empty. filterMiss.recovered=false with unfilteredCount 0 is the only genuinely empty window.
get_reviews_in_range
ChatGPTIndividual reviews in a date range with optional rating bounds. Returns text, rating, date, language, consumer name, and companyReply (the business's own response to the review, null when none). Cap 50, newest first. Review tracking needs the Basic plan or higher.
get_tone_evolution
ChatGPTHistory of generated qualitative profiles for a competitor, newest first. Cap 6 profiles.
get_top_ads_across_period
ChatGPTTop ad creatives across a multi-month window from digest top10Ads (parallel to get_top_posts_across_period). sortBy: euReach (default for facebookAds), recency (default for googleAds), duration (longest-running first). Cap 30. Ad tracking needs the Basic plan or higher.
get_top_posts_across_period
ChatGPTBest-performing posts across a multi-month window by engagement (likes + comments + shares). Months are YYYY-MM. Cap 30.
get_top_reviews_across_period
ChatGPTTop reviews across a multi-month window from digest top10Reviews. sortBy: rating_desc (praise patterns, default), rating_asc (complaint patterns), recency. Each review includes companyReply (the business's own response, null when none). Cap 30. Review tracking needs the Basic plan or higher.
get_web_analytics
ChatGPTThe brand's OWN website from its connected Google Analytics (GA4): visits (sessions), users, key events (what the business counts as a conversion; null = none set up), revenue when the site reports it, visits by channel group, the top UTM campaigns, and which of the brand's published posts brought visits through their tracked links (7 and 30 days after posting). Default window 28 days, 7-90 allowed; it ends on the last day Google Analytics finished counting. Returns { connected:false } when GA4 is not connected for this brand. Counts are a floor (cookie consent). Null means NOT MEASURED, never zero. Google Analytics needs the Starter plan or higher.
list_brands
ChatGPTList the brands this user can reach (their own account's brands plus any shared with them as a guest) and which one THIS connection reads. Read-only and free: it cannot switch brands — a brand is chosen per connection, and the answer's howToSwitch says how for this connection. Call it when the user mentions a brand other than the active one.
list_own_ad_accounts
ChatGPTList the user's OWN connected paid-media ad accounts (Meta / Google / TikTok / etc., via the Studio ads integration) — NOT competitor ads. Use this to check whether ad data is available before pulling performance. Returns an empty list when no ad account is connected.
list_own_ad_campaigns
ChatGPTList the user's OWN paid ad campaigns (NOT competitor ads) with compact per-campaign metrics (name, platform, status, spend, ROAS, conversions, clicks). Use for a campaign-by-campaign breakdown. Optional status filter and date window. Cap 50.
search_competitors
ChatGPTFind the user's tracked competitors matching a query string. Matches on name and platform handles (case- and diacritic-insensitive). Returns up to 20 results, each with links — the website, social profile, Trustpilot and ad-library URLs the competitor is tracked with (null = not configured). Use it to answer "what is their Instagram / website URL?".