Ayzeo

Generates a new content page for the given query using the N8N content-page pipeline, persists it, and returns the ContentPage id + generated SEO elements + edit URL. Counts against the project apiLimiter quota. Requires an active paid subscription on the project owner.

create_content_page_from_query

Returns Google Analytics 4 traffic segmented by AI referrer source (ChatGPT, Claude, Perplexity, Gemini, DeepSeek, Grok, Copilot, You.com, Phind, Kagi, OpenAI API apps) plus organic/direct/social baselines. Per source (bySource): sessions, new users, engagement rate, bounce rate, avg session duration, key events (conversions = all GA key events summed, an event count, not converting sessions), the project's chosen key event (primaryKeyEvents), ecommerce purchases, purchase revenue (in currencyCode) and add-to-carts; topLandingPages lists the pages AI visitors land on. paidAI covers ChatGPT Ads separately (never counted in totalAISessions or aiTrafficPercentage): ads vs organic ChatGPT outcomes, a daily split, the campaign table (utm_campaign / utm_id / utm_content) and untaggedClicks, i.e. ad clicks that arrived without UTMs and are credited to other sources (an estimate). trafficSpikes lists days where an organic AI source looks like bot traffic: reason "burst" (far more sessions than its usual day, mostly bounced) or "engagement_drop" (more sessions and a bounce rate several times its usual one). sampledDates lists days GA answered from a sample of the traffic (estimates; recent ones are re-fetched by later syncs). All rates (engagementRate, bounceRate, baselineBounceRate) are percentages 0-100. When the detailed read fails, bySource falls back to sessions, new users, conversions, purchases and revenue only. Requires GA integration. No distinct-user (unique visitor) count is available: GA4 reports it per dimension row and it cannot be summed over a date range, so only additive metrics are returned.

get_ga_ai_traffic

Returns Ayzeo's curated GEO playbook: recommended tactics for improving brand visibility in AI search engines, maintained as per-model sets (ChatGPT, Claude, Perplexity, Grok; Gemini/AI Overviews are covered by classic SEO). Pass `engine` to get that engine's resolved set: engine-specific texts and priorities merged in, sorted from high to low priority, each recommendation stamped with last_updated and the model version it was validated against. Omit `engine` for the full playbook (base recommendations plus per_engine override blocks and the per-engine model-version map) so the calling agent can reason across engines itself.

get_geo_recommendations

Headline content-prioritization tool. Returns the top-N highest-ACOS (Ayzeo Content Opportunity Score) GSC queries for the project. ACOS is a 0–100 deterministic composite score combining 9 weighted dimensions: strikingDistance (0.20), ctrGap (0.10), impressionVolume (0.15), momentum (0.10), positionTrend (0.05), contentGap (0.15), brand (0.10), longTail (0.05), trafficUplift (0.10). The tool surfaces the per-dimension scores, weights, and underlying GSC metrics (impressions, position, trend deltas, topPage) so the AI can articulate its own reasoning. Each item also flags hasDedicatedContentPage / existingContentPageId for clean branching between "create new" vs "enhance existing". Requires GSC integration to be connected and synced. Implementation note: the underlying aggregation pulls the top 500 queries by 30-day impressions before scoring, so projects with >500 low-impression queries may miss some long-tail opportunities — tune minImpressions/maxPosition to re-shape the pre-filter if needed.

get_gsc_opportunities

Paginated raw aggregated GSC query data (unscored) for ad-hoc AI filtering. Sortable by impressions, clicks, ctr, or position. Requires GSC integration.

get_gsc_queries

Link health of the project's OWN-domain URLs that AI answers exposed, the same data as the URL Health page. layer "answers" (default) = links shown inside the answer text, the ones a reader sees and clicks; layer "all" adds pages the models opened or consulted while answering (machine-side, never shown to a reader). Each URL carries the measured status: ok (200), redirect (finalUrl + redirectChain), dead (404/410), soft404 (the page answers but its content says not found, evidence-based), unconfirmed (missed one check, being rechecked, NOT broken yet), unreachable (timeout, server error, blocked by the site or robots.txt, NEVER counted as broken) or pending (not checked yet). broken is true only for dead and soft404, each confirmed by two separate checks at least 15 minutes apart. possibleFix is a pattern-based candidate that answered when checked (e.g. a dropped path segment, or the page's structure in the site's other markets): always present it as a possible fix to verify before redirecting, never as the confirmed fix; possibleFix.rule is a wildcard redirect when several broken URLs share one transformation. answers = distinct answers in the window that showed the URL in their text (layer "answers"), or that showed, opened or consulted it (layer "all"); prompts = distinct prompts, counted the same way; both count live (still tracked) prompts only. Sorted by answers, most-shown first. totals = per-status counts of every URL in the window and layer (not only the returned rows); total = rows matching `status` before `limit`. status: "dead" = broken only, "redirect", "unreachable", "ok", or "all" (default, includes unconfirmed and pending). hostsBlocked names hosts the latest checks could not check, with a reason (a refusal, timeouts, server errors, an unreadable robots.txt or a robots.txt Crawl-delay too long; their rows keep their last measured status), and the notes say which of them allowlisting the user agent fixes. Read-only: it never starts a check; checks run after the project's scans. Use for "which links does AI show to our site that are broken", "what should we redirect", and the broken-links section of a report.

get_url_health

Returns brand-mention data from AI answers aggregated per brand with sentiment counts (positive/neutral/negative) plus the 25 most recent individual mentions. Gives the AI a complete picture of the brand's positioning before recommending content.

get_brand_mentions

Returns recent prompt citation history across all AI models (ChatGPT, Claude, Perplexity, Gemini, DeepSeek, Grok) including citation counts, URL visits, mentionPosition (order the brand was named in that answer, 1 = first, null = not mentioned) with mentionedBrandCount, dominant sentiment, searchTerms (the web searches the AI ran for that answer — the query fan-out), citedUrls (pages on the project's own domain/channels the AI opened) and visitedUrls (every page the AI actually opened, competitors included). Each row carries promptDeleted: true when that prompt has since been removed from tracking (its history is kept); promptText is null only for runs whose prompt was hard-deleted before soft delete existed.

get_citation_history

Aggregates the sources behind the AI answers: which domains the AI models' web searches surfaced across ALL the project's prompts in the window (latest run per prompt+model). Returns up to 60 domains ranked by citations, each with class ('own' = the project's own domains, 'pinned' = a tracked competitor, 'other' = third-party such as directories, review sites, press — the quickest listing wins), citations, prompts (distinct prompts it appeared for), answers (distinct answers whose search results included it) and answerShare (answers as a % of answersTotal, the number of latest answers in the window), models, and sourceType where classified, plus a mix summary, totalCitations and answersTotal. Pass `domain` to drill into one domain: its most-cited URLs (title, citations, answers, answerShare, path bucket), the detail's own answersTotal denominator, and the prompts driving them. Pass `view: "pages"` for the page-level list instead: URLs across every domain ranked by citations (default 100 rows, `limit` up to 300; fields url, domain, title, class, sourceType, citations, prompts, answers, answerShare, models, bucket, unresolved for opaque Google redirect links) plus totalUrls (distinct URLs before the cap) and a `buckets` page-type mix computed over the full cached top-300 list, not only the rows returned under `limit`. Pass `url` to drill into one exact page: its citations, answerShare and the prompts whose answers consulted it (takes precedence over `domain`). Windowed (default 30d; presets up to "all", or startDate+endDate). Use for "which sources do AI answers rely on", "where can I get listed", and content-gap analysis against competitor pages.

get_citation_sources

Returns the project's already-generated content pages with their query, template, enhancement state, placement status, SEO HTML, and edit URLs. Use before recommending a new page to check whether one already exists.

get_content_pages