These are the workflows MarkMate actually ships. Every card is generated from its executable JSON and cross-checked against generally available registry operations.
Check one critical URL and, after two completed probes return unhealthy HTTP results, prepare an external webhook alert. The alert never sends automatically: MarkMate pauses until a human approves the exact destination and payload. DNS, TLS, connection, and timeout errors fail the workflow before the alert node and remain visible as failed-run history signals; this template does not convert those transport failures into webhook alerts.
Check URL statusweb_check_url_status
Starts immediately
Check conditioncondition
After first_probe
Wait 15 secondswait
After first_failed
Check URL statusweb_check_url_status
After pause
Check conditioncondition
After confirm_probe
Send webhook (POST)web_post_webhook
Approval After first_probe + confirm_probe + confirmed_failed
Audit a selected page sample, reduce the findings, and ask the configured AI model to propose a bounded technical priority list with acceptance checks.
Turn a bounded brand kit and campaign objective into reviewable ad copy and visual-direction variants. The AI step creates drafts only; it never publishes, spends, or approves a write.
Generate ad creative variantscreative_generate_ad_variants
Fetch your page and up to 10 competitor pages, compare term profiles deterministically (topics a majority of competitors cover that you don't, shared topics, and your unique angles), then have the AI write a prioritized plan to close the gaps.
Capture your page and a bounded competitor page set, reduce their observable term differences, and preserve the raw page evidence with a deterministic dossier.
Compare one owned page with selected competitor pages across readable copy, technical page quality, topic coverage, calls to action, trust signals, and structured-data readiness, then produce an evidence-backed differentiation and experiment plan.
Fetch page textweb_fetch_page
Starts immediately
Audit page SEOweb_audit_page
Starts immediately
Fetch page textweb_fetch_page
Loop Starts immediately
Audit page SEOweb_audit_page
Loop Starts immediately
Content gap vs competitorsseo_content_gap
After own_copy + rival_copy
Analyze data with AIai_analyze_data
After own_copy + own_audit + rival_copy + rival_audit + gap
Compare one page with a bounded competitor sample, then ask the configured AI model for an evidence-led content brief while preserving the deterministic gap result.
Compare one page with a bounded competitor sample, pair the deterministic term gap with a technical audit of your page, and save one evidence-led triage report.
Sample a site's sitemap, audit each selected page, aggregate repeated technical failures, and produce a crawl-efficiency priority map focused on indexability, canonical consistency, redirects, thin pages, and repeated template-level defects. A guard step stops the run if the URL is a sitemap index, so point the workflow at a child sitemap that lists page URLs directly.
Sample a site's priority inventory, audit technical SEO, conversion, and AI-visibility signals, aggregate recurring issues, and render an executive scorecard that translates page-level evidence into business risk, decisions, owners, and a focused next-month agenda. A guard step stops the run if the URL is a sitemap index, so point the workflow at a child sitemap that lists page URLs directly.
Turn a bounded, reviewed FAQ into schema.org FAQPage JSON-LD and save the result as an auditable report artifact. It is a drafting aid only and makes no eligibility or search-result promise.
Read a FAQ page, generate an advisory FAQ JSON-LD draft from explicit facts, and ask the configured AI model to identify evidence and review risks before publication.
Collect separate GA4 campaign and landing-page views for the current period and produce a cautious investigation brief without pretending the two reports form a row-level causal join.
Run three bounded GA4 acquisition breakdowns to review campaign naming, source and medium coverage, engagement, key events, and revenue. The final brief distinguishes tracking-hygiene findings from performance observations and does not assign a synthetic campaign quality score.
Review current-period GA4 channel-group rows for observed traffic, engagement, key-event, and revenue patterns, then produce a measurement-safe quality brief without assigning channel causation.
Review current-period GA4 country aggregates for observed traffic, engagement, key events, and revenue signals, then create localization questions without treating country as a causal or market-fit conclusion.
Compare the latest seven complete days with the preceding seven days by GA4 session default channel group, then produce a concise, source-grounded brief of material traffic, engagement, conversion, and revenue changes. The workflow reports observed movement without claiming attribution or causes the data cannot establish.
Review current-period GA4 device-category aggregates for sessions, engagement, key events, and revenue, then create a responsive and instrumentation follow-up brief without device causation claims.
Review current-period GA4 event-name aggregates for volume, engagement, key-event, and revenue signals, then identify event taxonomy and instrumentation checks without certifying tracking quality.
Read daily GA4 engagement signals across the last 28 complete days and turn them into a compact trend brief. The analysis separates sustained movement from isolated daily variation, reports exact observed values, and avoids unsupported explanations for changes.
Review a bounded set of GA4 landing-page rows for high-traffic pages with weak engagement or conversion signals, then create an investigation brief. Results are diagnostic candidates from observed analytics data, not proof that a page caused lost key events or revenue.
Review a bounded GA4 landing-page report for observed traffic, engagement, key-event, and revenue signals, then produce a measurement-safe page investigation brief without assigning lost value or causal blame.
Review current-period GA4 page-path rows for sessions, key events, and recorded revenue, then produce a verification brief that avoids assigning revenue causation to pages.
Review current-period GA4 source and medium rows for naming hygiene, engagement, key-event, and revenue signals, keeping observed analytics separate from attribution claims.
Check a curated set of landing, checkout, signup, and campaign URLs, then turn completed HTTP status, redirect, and final-destination results into a concise triage report. This read-only workflow creates a private MarkMate report and does not notify external systems. DNS, TLS, connection, or timeout errors fail the workflow and appear in run history instead of producing a report.
Read a curated group of related pages, compare their topics and intent, and create a source-to-destination internal-link plan with suggested contextual anchors, placement rationale, and safeguards against repetitive or misleading links.
Cluster a bounded keyword list, ask the configured AI model to explain the evidence and content implications, and save the structured cluster data plus narrative brief in a report.
Combine deterministic keyword clustering with a small on-page audit sample and save one report that separates topic structure from measured page signals. No ranking or demand data is invented.
Cluster a supplied keyword set with deterministic token similarity and store the clusters, singletons, and coverage notes in a report. This is a transparent planning reducer, not a live rank or demand estimate.
Group a keyword list into topic clusters (deterministic token-similarity clustering — no external SEO API needed), then have the AI turn the clusters into a content plan: one article per cluster with a working title and target intent.
Audit one high-value landing page and save a compact, evidence-led scorecard covering technical SEO, calls to action, forms, page weight proxies, structured data, and social metadata. The workflow reads and reports; it never edits or publishes the page.
Run a preflight review of one launch or campaign page covering reachability, redirects, technical SEO, mobile and conversion proxies, readable offer copy, metadata, social sharing signals, and AI-search structured-data readiness.
Sample pages from a sitemap, audit their title, meta description, canonical, robots, H1, language, Open Graph, and structured-data signals, then produce a remediation inventory organized by repeated template defect and search-snippet risk. A guard step stops the run if the URL is a sitemap index, so point the workflow at a child sitemap that lists page URLs directly.
Combine the month's strategic keyword themes with live copy and technical evidence from priority pages, then create an editorial operations brief covering maintenance debt, cluster coverage, publishing opportunities, quality controls, dependencies, and a bounded next-month backlog.
Place GA4 Organic Search rows beside daily Search Console performance for the same 28-day window and produce an executive scorecard. The workflow highlights directional patterns while keeping platform definitions separate and refusing to present correlation as attribution.
Generate a validated Organization schema draft for a reviewed marketing site, then preserve the exact JSON-LD and its review warnings in a workspace report. The workflow never injects or deploys markup.
Audit a page's measurable HTML signals, generate a separate reviewed schema draft, and store both outputs in one report. The draft never assumes the page is ready to publish structured data.
Use one bounded on-page audit to identify observable CTA, form, offer, content, and page-weight friction signals for a human conversion review without inventing a conversion rate or declaring a cause.
Read one bounded page snapshot and ask the configured AI model to identify clarity, evidence, and review questions without pretending to measure performance.
Audit a curated acquisition-to-conversion journey and compact its measured page findings before prioritizing mobile viewport, page-weight, form-friction, call-to-action, metadata, and structured-data risks. This workflow uses code-level proxies rather than browser or field-performance evidence.
Check a launch URL's HTTP response, technical page signals, and bounded readable copy, then preserve the evidence in one report before a campaign goes live.
Audit one public page and inspect its readable content to identify heading hierarchy, link-context, navigation, and page-structure questions for a human SEO and content review.
Audit one public page and pair its structured metadata, headings, copy signals, and social preview fields with a bounded AI review. The result is a diagnostic checklist, not a browser, accessibility, or ranking certification.
Read a bounded page snapshot, generate an advisory FAQ schema draft from supplied facts, and preserve both the source evidence and draft warnings in a report.
Read one public page, preserve its title and bounded text evidence, and store a searchable report artifact for review. It is useful for a campaign brief or competitor teardown without adding unverified interpretation.
Combine HTTP status, on-page audit, and bounded page evidence to review script count, response signals, and page-weight risks before a campaign or release.
Inspect one page's audit and readable source-derived text to review structured-data evidence, canonical context, headings, and social metadata with explicit limits around eligibility and search display.
Read a pricing page, generate an advisory Product JSON-LD draft from explicit example facts, and save the page evidence plus schema warnings for human review.
Probe a bounded set of selected URLs, preserve the final destinations and redirect flags, and write a concise chain-evidence report for technical review.
Probe a migration map of legacy URLs, capture each observed status, redirect flag, and final destination, then compare the results with expected targets to identify mismatches, broken sources, avoidable redirect risk, and cases needing hop-by-hop verification.
Fetch a site's robots.txt and check the homepage status so an AI reviewer can summarize visible crawl directives, host reachability, and follow-up questions without pretending to model the full crawler or index.
Use explicit brand terms to produce a model-assisted narrative grouping of a bounded Search Console query sample. The report keeps ambiguous examples visible and does not claim exhaustive classification, row preservation, total demand coverage, or market share.
Review a bounded Search Console country-device aggregate to surface observed market and responsive-search questions while keeping sparse combinations, omitted rows, and causation explicit.
Combine country-level and country-by-query Search Console views into a geographic demand brief for the returned sample. Country keys and metrics are preserved as reported, while the analysis avoids market-size estimates, location intent assumptions, and unsupported expansion claims.
Review current-period Search Console country aggregates to identify where search demand is observed and which localization or market checks deserve validation, without treating volume as revenue or market causation.
Summarize current-period Search Console device rows and surface observed click, impression, CTR, and position differences for a responsive-search review without claiming a device caused performance changes.
Compare Search Console performance first by device and then by device plus query to find returned queries whose clicks, CTR, or position differ materially across device categories. The workflow treats those differences as investigation opportunities, not proof of a device-specific experience problem.
Review current-period Search Console page aggregates to identify pages with observed clicks, impressions, CTR, and position signals that deserve indexing and content verification.
Review a bounded Search Console page and query sample to identify observed intent clusters and pages worth investigation, while keeping query evidence, page evidence, and omitted rows explicit.
Review a bounded current-period Search Console page sample and surface pages with weak CTR, attainable average positions, or meaningful click demand worth protecting. This snapshot does not measure decay or sustained deterioration because it does not query a comparable prior period, and it does not represent every indexed page.
Review current-period Search Console page and search-appearance aggregates to identify pages and features needing markup or snippet validation without claiming eligibility or complete visibility.
Review current-period Search Console query aggregates for impression, click, CTR, and position opportunities, with explicit sample limits and no unsupported momentum or ranking claims.
Review a bounded current-period Search Console query sample and surface high-impression, weak-CTR, and attainable-position opportunities for investigation. This snapshot does not measure momentum, decline, or newly visible demand because it does not query a comparable prior period, and omitted rows remain unknown.
Review current-period Search Console search-appearance aggregates to surface observed rich-result and feature visibility questions, with exact evidence and clear limits around coverage, eligibility, and causation.
Fetch your sitemap, run an on-page technical-SEO audit of a sample of its pages (title, meta description, canonical, indexability, H1s, mobile viewport, image alt coverage, content length, Open Graph — scored 0–100), aggregate the sampled pages into one scorecard, and have the AI write a prioritized action plan. A guard step stops the run if the URL is a sitemap index, so point the workflow at a child sitemap that lists page URLs directly.
Fetch your sitemap, run an on-page technical-SEO audit of a sample of its pages (title, meta description, canonical, indexability, H1s, mobile viewport, image alt coverage, content length, Open Graph — scored 0–100), aggregate the sampled pages into a scorecard, have the AI write a prioritized action plan, and render it all into a branded HTML report saved to Files (open it and print to PDF, or attach it to a client). A guard step stops the run if the URL is a sitemap index, so point the workflow at a child sitemap that lists page URLs directly. This is the reporting workflow MarkMate runs on itself.
Reduce supplied organic SERP observations into rank-weighted share of voice, then produce an advisory brief that keeps observed visibility separate from causes and business outcomes.
Measure competitor SERP share of voiceseo_competitor_share_of_voice
Calculate transparent rank-weighted share of voice from supplied organic observations and save the domain summaries and methodology warnings. It does not fetch SERPs, estimate traffic, or claim causality.
Measure competitor SERP share of voiceseo_competitor_share_of_voice
Check a site root and read its sitemap discovery response to create a compact endpoint report, keeping HTTP reachability, sitemap indexing, and crawl coverage as separate facts.
An autopilot rule: audit a fixed sample of pages from your sitemap on a schedule, and when the sample's average SEO score drops below your threshold, send an alert webhook (Slack/Discord/Zapier) — after you approve it. A guard step stops the run if the URL is a sitemap index, so point the workflow at a child sitemap that lists page URLs directly. The alert step is a write, so it always pauses in the approval inbox; if the score is healthy, the run ends quietly.
Sample page URLs from a URL-set sitemap, fetch bounded text, and review visible language, titles, headings, and content consistency without claiming complete localization or content quality coverage.
Get sitemap URLsweb_get_sitemap
Starts immediately
Check conditioncondition
After sitemap
Take first/last N of a listutil_take
After sitemap + urlset_guard
Fetch page textweb_fetch_page
Loop After sample
Analyze data with AIai_analyze_data
After sample + pages
Render an HTML report artifactreport_render
After sitemap + urlset_guard + sample + pages + analysis
Read a bounded sitemap page sample, ask the configured AI model to organize observable content themes and review questions, and save the source sample with the advisory brief.
Fetch a bounded sample of pages from a URL-set sitemap and review titles, visible text, headings, and readability signals while preserving the difference between sampled evidence and a complete editorial audit.
Get sitemap URLsweb_get_sitemap
Starts immediately
Check conditioncondition
After sitemap
Take first/last N of a listutil_take
After sitemap + urlset_guard
Fetch page textweb_fetch_page
Loop After sample
Analyze data with AIai_analyze_data
After sample + pages
Render an HTML report artifactreport_render
After sitemap + urlset_guard + sample + pages + analysis
Inspect one sitemap or sitemap index and preserve its URL coverage, truncation, and index-shape evidence in a report. The workflow does not crawl or publish anything.
Read a URL-set sitemap, sample its newest entries, check their HTTP status, and produce a bounded freshness-risk brief. It reports current responses only and does not infer publication dates or full sitemap coverage.
Get sitemap URLsweb_get_sitemap
Starts immediately
Check conditioncondition
After sitemap
Take first/last N of a listutil_take
After sitemap + urlset_guard
Check URL statusweb_check_url_status
Loop After sample
Analyze data with AIai_analyze_data
After sample + probe
Render an HTML report artifactreport_render
After sitemap + urlset_guard + sample + probe + analysis
Sample URLs from a child sitemap, probe their final HTTP destinations, and store a bounded redirect and availability report. This is useful during releases and migrations without pretending to measure every redirect hop.
Sample a URL-set sitemap and check redirect behavior for a bounded set of pages, producing a release-oriented report that keeps HTTP evidence separate from canonical and SEO conclusions.
Get sitemap URLsweb_get_sitemap
Starts immediately
Check conditioncondition
After sitemap
Take first/last N of a listutil_take
After sitemap + urlset_guard
Check URL statusweb_check_url_status
Loop After sample
Analyze data with AIai_analyze_data
After sample + redirects
Render an HTML report artifactreport_render
After sitemap + urlset_guard + sample + redirects + analysis
Read a production sitemap, sample a configurable slice of its URLs, verify that they resolve successfully, and create a release-oriented report covering missing inventory, accidental redirects, and pages that need immediate search visibility review. A guard step stops the run if the URL is a sitemap index, so point the guard at the child sitemap for the released surface.
Audit a bounded sample of sitemap pages, aggregate technical checks into a scorecard, and save a reviewable report without generating a causal or ranking claim.
Sample pages from a URL-set sitemap, inspect their on-page structured-data signals, and produce a cautious AI review of schema coverage and validation follow-ups.
Get sitemap URLsweb_get_sitemap
Starts immediately
Check conditioncondition
After sitemap
Take first/last N of a listutil_take
After sitemap + urlset_guard
Audit page SEOweb_audit_page
Loop After sample
Analyze data with AIai_analyze_data
After sample + audits
Render an HTML report artifactreport_render
After sitemap + urlset_guard + sample + audits + analysis
Check a bounded tail sample from a URL-set sitemap and create a direct availability watch report for recently listed entries, without interpreting sitemap order as publication time.
Get sitemap URLsweb_get_sitemap
Starts immediately
Check conditioncondition
After sitemap
Take first/last N of a listutil_take
After sitemap + urlset_guard
Check URL statusweb_check_url_status
Loop After sample
Analyze data with AIai_analyze_data
After sample + statuses
Render an HTML report artifactreport_render
After sitemap + urlset_guard + sample + statuses + analysis
Sample a URL-set sitemap, run bounded on-page technical audits, aggregate findings, and render a site audit artifact with explicit sample limits and no synthetic full-site score.
Get sitemap URLsweb_get_sitemap
Starts immediately
Check conditioncondition
After sitemap
Take first/last N of a listutil_take
After sitemap + urlset_guard
Audit page SEOweb_audit_page
Loop After sample
Summarize SEO auditsweb_audit_summary
After audits
Render an HTML report artifactreport_render
After sitemap + urlset_guard + sample + audits + summary
Read a URL-set sitemap, guard against sitemap indexes, check a bounded set of listed URLs, and render direct HTTP status and redirect evidence without claiming full-site uptime.
Audit a curated portfolio of evergreen or commercial pages, compact the measured findings, and rank technical refresh candidates using metadata, indexability, structured-data, content-depth, and conversion-friction signals. The workflow does not claim copy freshness, publication age, traffic decline, or factual accuracy.
Compare bounded GA4 landing-page traffic rows with Search Console page visibility for the same period, then surface page-level mismatches worth checking. The workflow explicitly treats GA4 sessions and Search Console clicks as different measurements and does not manufacture a cross-platform attribution model.
Select GA4 propertiesga4_select_accounts
Starts immediately
Select Search Console sitesgsc_select_accounts
Starts immediately
Resolve a date rangeutil_date_range
Starts immediately
Get GA4 reportga4_get_report
After property + range
Get Search Console reportgsc_get_report
After site + range
Check conditioncondition
After traffic_pages
Check conditioncondition
After search_pages
Analyze data with AIai_analyze_data
After traffic_pages + search_pages + has_traffic + has_search
Check one URL's HTTP destination, readable content, and technical audit together, then produce a compact dossier that separates transport, page, and editorial evidence.
Collect separate GA4 channel, source-medium, and campaign views plus a bounded Search Console query-page view for the last seven complete days. The final report summarizes acquisition mix and search demand while keeping channel performance, search visibility, and attribution claims distinct.
Select GA4 propertiesga4_select_accounts
Starts immediately
Select Search Console sitesgsc_select_accounts
Starts immediately
Resolve a date rangeutil_date_range
Starts immediately
Get GA4 reportga4_get_report
After property + range
Get GA4 reportga4_get_report
After property + range
Get GA4 reportga4_get_report
After property + range
Get Search Console reportgsc_get_report
After site + range
Analyze data with AIai_analyze_data
After channels + sources + campaigns + search
Render an HTML report artifactreport_render
After channels + sources + campaigns + search + analysis
Plan
Starter manual / Growth scheduled
Connection
Google Analytics 4, Google Search Console
Approval
No approval steps
Start from something executable
Open a workspace. Keep every requirement visible.
Free workspaces can inspect and build workflows. The catalog marks paid execution gates before you start.