30 Claude Prompts That Build SEO Dashboards

You are an SEO analyst who builds Search Console performance overviews a weekly SEO standup can defend. <context> I need a Search Console clicks overview as one self-contained HTML file with inline CSS, vanilla JS, and a CDN chart library, seeded with realistic daily performance so I can click the working screen in the chat. </context> <inputs> - Property: [PROPERTY] - Period: [E.G. LAST 90 DAYS] - Compare to: [E.G. PRIOR 90 DAYS] - Country or device filter if useful: [OR "ALL"] - Click target if any: [E.G. +10% OR "NONE"] </inputs> <task> Build the dashboard with: a header for [PROPERTY]; KPI cards (clicks, impressions, average CTR, average position, versus prior); a dual-axis line of clicks and impressions; a CTR line with the period average; a weekday table (day, clicks, impressions, CTR); and a latest-week callout versus the prior four weeks. Keep a daily series in labeled JS constants and compute CTR as clicks / impressions. Versus-prior uses matching-length windows. Do not invent conversions. Search metrics only. </task> <constraints> - One self-contained HTML file; one CDN charting library only. - Daily rows in labeled JS constants at the top so I can swap a Search Console performance export later. - Do not invent a live API. Sample data only. - Below-average CTR weeks flagged. Accessible contrast. </constraints> <format> Return the full HTML so I can click the working screen in the chat, then write the CTR and versus-prior math and which series to replace. </format>

You are an SEO analyst who finds title and meta work by comparing actual CTR to a position-based expected CTR. <context> I need a CTR versus position dashboard as one self-contained HTML file with inline CSS, vanilla JS, and a CDN chart library, seeded with realistic queries so I can click the working screen in the chat. </context> <inputs> - Site: [SITE] - Period: [E.G. LAST 28 DAYS] - Expected CTR by position band: [OR "USE A SIMPLE EDITABLE TABLE"] - Minimum impressions to score: [E.G. 200] - Queries: [OR "INVENT REALISTIC ONES"] </inputs> <task> Build the dashboard with: a header for [SITE]; KPI cards (queries scored, median CTR gap, title-fix count, skipped as thin); a scatter of position versus CTR with the expected curve; a gap table (query, position, impressions, actual CTR, expected CTR, gap); and a title-fix list where actual sits under expected by an editable threshold. Compute actual CTR from clicks / impressions. Expected CTR comes from an editable band table in the constants (for example position 1 = 28 percent, 2 = 15, 3 = 11, 4-10 sliding down). Queries under the impression floor show as thin, not as a wild gap. Clicking a gap row highlights the scatter point. </task> <constraints> - One self-contained HTML file; one CDN charting library only. - Queries and the expected-CTR table in labeled JS constants at the top. - Show the expected table on the page. Do not invent a proprietary CTR model. - Accessible contrast. Thin rows labeled thin. </constraints> <format> Return the full HTML so I can click the working screen in the chat, then write the gap formula and which constants to replace with a query CTR table. </format>

You are an SEO strategist who turns high-impression, low-click queries into a rewrite board, not a keyword fantasy list. <context> I need a query click-opportunity dashboard as one self-contained HTML file with inline CSS, vanilla JS, and a CDN chart library, seeded with realistic Search Console style queries so I can click the working screen in the chat. </context> <inputs> - Query set: [QUERY SET] - Site: [SITE] - Period: [E.G. LAST 28 DAYS] - Opportunity rule: [E.G. IMPRESSIONS OVER 500 AND CTR UNDER 2% AND POSITION UNDER 15] - Queries: [OR "INVENT PLAUSIBLE ONES"] </inputs> <task> Build the dashboard with: a header for [QUERY SET]; KPI cards (queries, opportunity rows, wasted impressions, estimated missed clicks); a bar of wasted impressions by query; a sortable table (query, impressions, clicks, CTR, position, wasted impressions, suggested action title/meta or content); and a rewrite list. Compute wasted impressions as impressions minus clicks. Estimate missed clicks as impressions times (expected CTR minus actual CTR) using a simple editable expected CTR, and mark the estimate as an estimate. Rows that fail my opportunity rule stay in the full table but not on the rewrite list. Clicking a row shows the rule check. </task> <constraints> - One self-contained HTML file; one CDN charting library only. - Query rows and the rule in labeled JS constants at the top so I can swap a Search Console query export later. - Do not invent rankings you did not seed. Sample data only. - Accessible contrast. Sort default: wasted impressions descending. </constraints> <format> Return the full HTML so I can click the working screen in the chat, then write wasted-impression math and which array to replace. </format>

You are an SEO analyst who reads one landing page as a weekly click trend, not a lifetime total. <context> I need a page click-trend dashboard as one self-contained HTML file with inline CSS, vanilla JS, and a CDN chart library, seeded with a realistic weekly series so I can click the working screen in the chat. </context> <inputs> - Landing page path: [LANDING PAGE] - Site: [SITE] - Weeks: [E.G. LAST 16] - Known changes if any: [E.G. TITLE REWRITE IN WEEK 9 OR "NONE"] - Drop rule: [E.G. CLICKS DOWN 25% VS PRIOR 4-WEEK AVERAGE] </inputs> <task> Build the dashboard with: a header for [LANDING PAGE]; KPI cards (latest week clicks, CTR, position, versus prior 4 weeks); a dual line of weekly clicks and impressions; a CTR line; a table of weeks; markers on the chart for [E.G. TITLE REWRITE IN WEEK 9 OR "NONE"]; and a drop callout using my rule. Compute CTR each week from that week's clicks and impressions. A change marker is a date I gave, not a claimed cause. Do not write that the rewrite caused the drop. Timing only. Clicking a week highlights the table row. </task> <constraints> - One self-contained HTML file; one CDN charting library only. - Weekly series and markers in labeled JS constants at the top so I can swap that URL's Search Console performance later. - No causal language in labels. Accessible contrast. - Do not invent a second URL. </constraints> <format> Return the full HTML so I can click the working screen in the chat, then explain drop math and which series to replace. </format>

You are an SEO analyst who splits Search Console clicks by device and country a team can actually act on. <context> I need a device and country clicks dashboard as one self-contained HTML file with inline CSS, vanilla JS, and a CDN chart library, seeded with realistic dimension rows so I can click the working screen in the chat. </context> <inputs> - Property: [PROPERTY] - Period: [E.G. LAST 28 DAYS] - Devices: [MOBILE, DESKTOP, TABLET] - Countries to show: [6-10 OR "INVENT A REALISTIC MIX"] - Lagging rule: [E.G. CTR 30% BELOW PROPERTY CTR WITH 1K+ IMPRESSIONS] </inputs> <task> Build the dashboard with: a header for [PROPERTY]; a cut selector (device, country); KPI cards (clicks, impressions, CTR, lagging slices); a share bar; a CTR bar by slice; a table (slice, clicks, impressions, CTR, average position, versus property CTR); and a lagging-geo or lagging-device list using my rule. Compute CTR and shares from the dimension rows. Selecting a country filters the device breakdown for that country from a nested sample array, or shows n/a if I did not seed the nest. Slices under an editable impression floor show n/a CTR, not a wild score. </task> <constraints> - One self-contained HTML file; one CDN charting library only. - Dimension rows in labeled JS constants at the top so I can swap a Search Console device/country export later. - n/a labeled when thin. Accessible contrast. - Do not invent conversions or revenue. </constraints> <format> Return the full HTML so I can click the working screen in the chat, then explain the lagging rule and which arrays to replace. </format>

You are a technical SEO analyst who builds index coverage overviews from Search Console coverage style counts. <context> I need an index coverage overview as one self-contained HTML file with inline CSS, vanilla JS, and a CDN chart library, seeded with realistic coverage buckets so I can click the working screen in the chat. </context> <inputs> - Site: [SITE] - Snapshot date: [DATE] - Buckets: [INDEXED, EXCLUDED, ERROR, VALID WITH WARNING] - Counts: [OR "INVENT PLAUSIBLE"] - Prior snapshot if any: [OR "INVENT A SMALL SHIFT"] </inputs> <task> Build the dashboard with: a header for [SITE] and the snapshot date; KPI cards (indexed, excluded, error, change versus prior); a donut or stacked bar of buckets; a 6-week trend of indexed versus error if you seed a short series; a reason table (reason, count, share, severity); and an error-slice callout. Compute shares from the counts. Clicking a bucket filters the reason table. Severity is an editable map (error = high, excluded crawled currently not indexed = medium, and so on), shown on the page. </task> <constraints> - One self-contained HTML file; one CDN charting library only. - Bucket counts and reasons in labeled JS constants at the top so I can swap a coverage export later. - Do not invent a live Search Console API. Sample data only. - Error slice visually loud. Accessible contrast. </constraints> <format> Return the full HTML so I can click the working screen in the chat, then write share math and which constants to replace. </format>

You are a technical SEO analyst who builds excluded-URL watchlists that force a fix, noindex, or leave-it decision. <context> I need an excluded URL watchlist as one self-contained HTML file with inline CSS, vanilla JS, and a CDN chart library, seeded with realistic excluded rows so I can click the working screen in the chat. </context> <inputs> - Site: [SITE] - Reasons to include: [E.G. CRAWLED CURRENTLY NOT INDEXED, DISCOVERED NOT INDEXED, ALTERNATE WITH CANONICAL, BLOCKED BY ROBOTS, EXCLUDED BY NOINDEX] - Sample URLs: [12-20 OR "INVENT REALISTIC PATHS"] - Fix-first rule: [E.G. MONEY URL + CRAWLED NOT INDEXED + AGE OVER 14 DAYS] - Today constant: [DATE] </inputs> <task> Build the dashboard with: a header for [SITE]; KPI cards (excluded rows, crawled not indexed, discovered not indexed, fix-first count); a bar of count by reason; a sortable watchlist (url, reason, first seen, age days, page type if tagged, suggested action); reason chips that filter the table; and a fix-first list using my rule. Compute age from first seen and a today constant. Suggested actions come from an editable reason-to-action map (for example alternate with canonical = confirm canonical, noindex = leave or remove noindex). Clicking a row opens the suggested action. </task> <constraints> - One self-contained HTML file; one CDN charting library only. - URL rows and the action map in labeled JS constants at the top so I can swap an excluded-URL export later. - Suggested actions are rules I can edit, not a strategy memo. - Accessible contrast. Sort default: age descending. </constraints> <format> Return the full HTML so I can click the working screen in the chat, then write the age and fix-first rules and which array to replace. </format>

You are a technical SEO analyst who builds crawl error boards that separate broken cleanup from live incidents. <context> I need a crawl error dashboard as one self-contained HTML file with inline CSS, vanilla JS, and a CDN chart library, seeded with realistic status rows so I can click the working screen in the chat. </context> <inputs> - Property: [PROPERTY] - Crawl date: [DATE] - Error samples: [12-20 URLS OR "INVENT REALISTIC ONES"] - Statuses: [E.G. 404, 410, 500, 503, SOFT 404] - Incident rule: [E.G. 5XX ON A TAGGED MONEY URL] </inputs> <task> Build the dashboard with: a header for [PROPERTY]; KPI cards (error URLs, 4xx, 5xx, incidents); a stacked bar of status over a short week series if you seed one; a sortable table (url, status, page type, last seen, inbound links if given, severity); chips for 4xx / 5xx / all; and a broken-URL incident list using my rule. Compute counts from the rows. Severity is an editable map (5xx = high, 404 on money = high, 404 on old blog = low). Search filters the table. Clicking a status chip updates every widget. </task> <constraints> - One self-contained HTML file; one CDN charting library only. - Error rows in labeled JS constants at the top so I can swap a crawl error report later. - Do not invent server logs. Status and URL only. - Incident rows red. Accessible contrast. Search works. </constraints> <format> Return the full HTML so I can click the working screen in the chat, then write severity rules and which array to replace. </format>

You are a technical SEO analyst who reconciles sitemap submitted URLs against what is indexed. <context> I need a sitemap versus indexed dashboard as one self-contained HTML file with inline CSS, vanilla JS, and a CDN chart library, seeded with realistic URL rows so I can click the working screen in the chat. </context> <inputs> - Sitemap name or URL: [SITEMAP] - Site: [SITE] - Submitted count: [OR "INVENT"] - Indexed count: [OR "INVENT"] - Sample missing and extra URLs: [OR "INVENT REALISTIC PATHS"] </inputs> <task> Build the dashboard with: a header for [SITEMAP]; KPI cards (submitted, indexed, missing from index, indexed but not in sitemap); a two-circle or grouped bar of submitted versus indexed; a missing-URL list (in sitemap, not indexed); an extra-URL list (indexed, not in sitemap); and a table with a status chip per URL (in both / missing / extra). Compute missing as submitted minus overlap, extra as indexed minus overlap, from a sample URL array with two boolean flags (inSitemap, indexed). Clicking missing filters the table. Do not invent lastmod facts I did not seed. Empty lastmod is allowed. </task> <constraints> - One self-contained HTML file; one CDN charting library only. - URL flags in labeled JS constants at the top so I can swap a sitemap status export later. - Show the overlap math on the page. Accessible contrast. - Do not invent a live sitemap API. </constraints> <format> Return the full HTML so I can click the working screen in the chat, then write overlap math and which array to replace. </format>

You are a technical SEO analyst who measures how long new URLs wait to be discovered and indexed. <context> I need a new-URL discovery lag dashboard as one self-contained HTML file with inline CSS, vanilla JS, and a CDN chart library, seeded with realistic new URLs so I can click the working screen in the chat. </context> <inputs> - Site: [SITE] - New URLs: [10-16 PATHS WITH PUBLISH DATES OR "INVENT"] - Discovery date and first index date per URL: [OR "INVENT CONSISTENT LAGS"] - Lag target days: [E.G. 7] - Today constant: [DATE] </inputs> <task> Build the dashboard with: a header for [SITE]; KPI cards (new URLs, median days to discovery, median days to index, still waiting); a histogram of lag days; a table (url, published, discovered, indexed, days to discovery, days to index, status); and a lag list where days to index exceed the target, or where index is still empty. Compute days from the dates. Still-waiting rows use the today constant and show as waiting, not as a fake zero lag. Clicking a status chip (indexed / waiting / discovered only) filters the table. </task> <constraints> - One self-contained HTML file; one CDN charting library only. - URL dates in labeled JS constants at the top so I can swap a URL inspection log later. - Waiting labeled waiting. Accessible contrast. - Do not invent rankings. Dates and lag only. </constraints> <format> Return the full HTML so I can click the working screen in the chat, then write lag math and which array to replace. </format>

You are an SEO analyst who finds queries where two or more of our URLs split impressions. <context> I need a query cannibalization finder as one self-contained HTML file with inline CSS, vanilla JS, and a CDN chart library, seeded with realistic query-by-page rows so I can click the working screen in the chat. </context> <inputs> - Focus query example: [QUERY] - Site: [SITE] - Period: [E.G. LAST 28 DAYS] - Cannibal rule: [E.G. TWO PLUS URLS EACH WITH 15%+ IMPRESSION SHARE ON THE SAME QUERY] - Sample query-page pairs: [OR "INVENT A REALISTIC SET"] </inputs> <task> Build the dashboard with: a header for [SITE]; KPI cards (queries scored, cannibalized queries, worst split, unique URLs in fights); a table of cannibalized queries (query, URL A, URL B, impressions A/B, clicks A/B, position A/B, impression share); a stacked bar of click share for the selected query; and a winner-pick note using a simple rule (most clicks, then best position) labeled as a rule, not a mandate. Compute impression share per URL on that query. Queries with only one URL do not appear on the fight list. Clicking a query updates the stacked bar and the winner note. </task> <constraints> - One self-contained HTML file; one CDN charting library only. - Query-page rows in labeled JS constants at the top so I can swap a Search Console query-by-page export later. - Show the cannibal rule on the page. Do not invent a third domain. - Accessible contrast. </constraints> <format> Return the full HTML so I can click the working screen in the chat, then write share and winner rules and which array to replace. </format>

You are an SEO analyst who clusters pages that share an exact title so the merge list is honest. <context> I need a duplicate-title cluster dashboard as one self-contained HTML file with inline CSS, vanilla JS, and a CDN chart library, seeded with realistic titles so I can click the working screen in the chat. </context> <inputs> - Site: [SITE] - Crawl of titles: [OR "INVENT 20-30 URLS WITH SOME DUPLICATE TITLES"] - Index status if known: [OR "UNKNOWN"] - Merge rule: [E.G. KEEP THE URL WITH MOST CLICKS] - Period for clicks if given: [E.G. LAST 28 DAYS] </inputs> <task> Build the dashboard with: a header for [SITE]; KPI cards (pages crawled, duplicate titles, clusters of 2+, singleton titles); a cluster list grouped by exact title; a selected-cluster table (url, title, clicks if given, index status); and a merge list that names a keep URL from my rule and the others as consolidate or retitle. Compute clusters in JS by grouping on normalized title (trim, collapse spaces, case fold). Singles do not appear on the merge list. Clicking a cluster opens its URLs. </task> <constraints> - One self-contained HTML file; one CDN charting library only. - URL-title rows in labeled JS constants at the top so I can swap a title crawl later. - Keep/consolidate is a rule, not a mandate. Accessible contrast. - Do not invent body copy. Titles and optional clicks only. </constraints> <format> Return the full HTML so I can click the working screen in the chat, then write the cluster and keep rules and which array to replace. </format>

You are an SEO strategist who finds pages that fight for the same intent because they share queries. <context> I need a same-intent page clash dashboard as one self-contained HTML file with inline CSS, vanilla JS, and a CDN chart library, seeded with overlapping query sets so I can click the working screen in the chat. </context> <inputs> - Topic: [TOPIC] - Candidate pages: [3-6 PATHS OR "INVENT"] - Top queries per page: [OR "INVENT OVERLAPS"] - Clash rule: [E.G. 3+ SHARED QUERIES OR 20%+ QUERY OVERLAP] - Site: [SITE] </inputs> <task> Build the dashboard with: a header for [TOPIC]; KPI cards (pages, clash pairs, strongest overlap, unique queries); a pair table (url A, url B, shared query count, overlap percent, sample shared queries); a Venn-style or two-column query list for the selected pair; and a keep-or-merge list using a simple rule (keep the URL with more clicks, merge or differentiate the other) labeled as a rule. Compute overlap as shared queries / union of queries for that pair. Pairs under the clash rule stay off the fight list. Clicking a pair updates the shared query list. </task> <constraints> - One self-contained HTML file; one CDN charting library only. - Page-query arrays in labeled JS constants at the top so I can swap overlapping query exports later. - Do not invent intent labels I did not give. Overlap math only. - Accessible contrast. </constraints> <format> Return the full HTML so I can click the working screen in the chat, then write overlap math and which arrays to replace. </format>

You are a technical SEO analyst who builds canonical conflict boards from declared and chosen URLs. <context> I need a canonical conflict dashboard as one self-contained HTML file with inline CSS, vanilla JS, and a CDN chart library, seeded with realistic crawl rows so I can click the working screen in the chat. </context> <inputs> - Site: [SITE] - Rows: [URL, DECLARED CANONICAL, CHOSEN CANONICAL OR "INVENT A MIX OF MATCH AND MISMATCH"] - Conflict types: [SELF OK, MISMATCH, CHAIN, LOOP, CROSS DOMAIN] - Money URL tags if any: [OR "NONE"] </inputs> <task> Build the dashboard with: a header for [SITE]; KPI cards (rows, mismatches, chains, loops); a type bar; a mismatch table (url, declared, chosen, type, money yes/no); chips for type; and a mismatch list with a suggested check (confirm declared, fix loop, drop cross-domain) from an editable map. Classify in JS: self ok when url equals declared and declared equals chosen or chosen is empty; mismatch when declared does not equal chosen; chain when declared points at a URL whose declared is a third URL; loop when A declares B and B declares A. Clicking a type filters the table. </task> <constraints> - One self-contained HTML file; one CDN charting library only. - Canonical rows in labeled JS constants at the top so I can swap a canonical crawl later. - Show the classification rules on the page. Accessible contrast. - Do not invent Google-chosen values I marked unknown. Unknown chosen is its own type. </constraints> <format> Return the full HTML so I can click the working screen in the chat, then write the classification rules and which array to replace. </format>

You are an SEO analyst who shows when internal links are split across twin URLs instead of concentrating on the money page. <context> I need an internal-link dilution dashboard as one self-contained HTML file with inline CSS, vanilla JS, and a CDN chart library, seeded with realistic inbound counts so I can click the working screen in the chat. </context> <inputs> - Money page: [MONEY PAGE] - Twin or competing URLs: [2-5 PATHS] - Unique inbound links per URL: [OR "INVENT PLAUSIBLE"] - Dilution rule: [E.G. A TWIN HAS 40%+ OF THE MONEY PAGE INBOUNDS] - Site: [SITE] </inputs> <task> Build the dashboard with: a header for [MONEY PAGE]; KPI cards (money inbounds, twin inbounds, dilution ratio, twins over the rule); a bar of unique inbounds by URL; a table (url, unique inbounds, share of the set, sample anchors if given); and a twin list that fails the dilution rule. Compute share as that URL's unique inbounds over the set total. Dilution ratio is the largest twin / money page. Clicking a twin lists sample linking paths from a small nested array. Do not invent PageRank. Counts only. </task> <constraints> - One self-contained HTML file; one CDN charting library only. - Inbound counts in labeled JS constants at the top so I can swap an internal-link export later. - Show the dilution rule on the page. Accessible contrast. - Unique linking URLs, not raw hits, labeled on the page. </constraints> <format> Return the full HTML so I can click the working screen in the chat, then write dilution math and which arrays to replace. </format>

You are a content SEO analyst who reads publishing pace next to organic clicks, not as a vanity post count. <context> I need a content velocity dashboard as one self-contained HTML file with inline CSS, vanilla JS, and a CDN chart library, seeded with a realistic publish log so I can click the working screen in the chat. </context> <inputs> - Blog or site: [BLOG] - Weeks: [E.G. LAST 16] - Target publishes per week: [E.G. 2] - Weekly organic clicks: [OR "INVENT A SERIES THAT LAGS PUBLISHES"] - Stall rule: [E.G. 0 PUBLISHES FOR 2+ WEEKS] </inputs> <task> Build the dashboard with: a header for [BLOG]; KPI cards (posts in range, average per week, versus target, latest-week clicks); a weekly bar of publishes with the target line; a line of organic clicks on the same weeks; a table (week, publishes, clicks, click change); and a stall-week callout using my rule. Keep publishes and clicks in separate labeled series. Do not claim a publish caused a click move. Show timing only. Clicking a stall week lists the nearby posts from a sample title array. </task> <constraints> - One self-contained HTML file; one CDN charting library only. - Weekly series in labeled JS constants at the top so I can swap a publish log and a Search Console weekly total later. - No causal language. Accessible contrast. - Below-target weeks flagged. </constraints> <format> Return the full HTML so I can click the working screen in the chat, then explain stall math and which series to replace. </format>

You are a content SEO analyst who measures how long a new URL waits to earn a first strong rank. <context> I need a publish-to-rank lag dashboard as one self-contained HTML file with inline CSS, vanilla JS, and a CDN chart library, seeded with realistic new URLs so I can click the working screen in the chat. </context> <inputs> - Content type: [CONTENT TYPE] - Site: [SITE] - New URLs with publish dates: [10-16 OR "INVENT"] - First date they hit a position at or above: [E.G. POSITION 10, OR "INVENT LAGS"] - Slow rule: [E.G. OVER 28 DAYS OR STILL WAITING] </inputs> <task> Build the dashboard with: a header for [CONTENT TYPE]; KPI cards (URLs, median days to page one, still waiting, slow count); a histogram of lag days; a table (url, published, first page-one date, days, current position, clicks); and a slow list. Compute days from publish to first page-one date. Still-waiting rows have no page-one date and use a today constant, shown as waiting. Page one means position at or under an editable threshold (default 10). Clicking a slow URL shows its sample weekly position sparkline. </task> <constraints> - One self-contained HTML file; one CDN charting library only. - URL dates and positions in labeled JS constants at the top so I can swap publish dates and a rank export later. - Waiting labeled waiting. Accessible contrast. - Do not invent traffic forecasts. </constraints> <format> Return the full HTML so I can click the working screen in the chat, then write lag math and which arrays to replace. </format>

You are a content SEO analyst who finds aging URLs that still earn impressions but are losing clicks. <context> I need a freshness decay dashboard as one self-contained HTML file with inline CSS, vanilla JS, and a CDN chart library, seeded with realistic page ages so I can click the working screen in the chat. </context> <inputs> - Site: [SITE] - Period versus prior: [E.G. LAST 28 VS PRIOR 28] - Pages: [12-20 WITH PUBLISH OR LAST-UPDATED DATES] - Decay rule: [E.G. AGE OVER 180 DAYS AND CLICKS DOWN 20% AND IMPRESSIONS STILL OVER 500] - Today constant: [DATE] </inputs> <task> Build the dashboard with: a header for [SITE]; KPI cards (pages scored, decaying, median age, prune-like count); a scatter of age versus click change; a sortable table (url, age days, last updated, clicks, prior clicks, change, impressions, flag decay/prune/ok); and a refresh list using my decay rule. Compute age from last updated or publish, using last updated when both exist. Prune-like means clicks and impressions both under editable floors. Clicking a decay row highlights the scatter point. Do not invent a refresh outline. Flags only. </task> <constraints> - One self-contained HTML file; one CDN charting library only. - Page rows in labeled JS constants at the top so I can swap ages and a Search Console pages export later. - Show decay and prune rules on the page. Accessible contrast. - Sort default: click change ascending. </constraints> <format> Return the full HTML so I can click the working screen in the chat, then write the flags and which array to replace. </format>

You are an SEO strategist who maps a pillar topic to supporting URLs and shows the holes. <context> I need a topic cluster coverage dashboard as one self-contained HTML file with inline CSS, vanilla JS, and a CDN chart library, seeded with a realistic cluster map so I can click the working screen in the chat. </context> <inputs> - Pillar topic: [PILLAR TOPIC] - Pillar URL: [PATH OR "INVENT"] - Planned subtopics: [8-12] - Existing URLs and their best query plus position: [OR "INVENT A PARTIAL MAP"] - Covered rule: [E.G. LIVE URL AND POSITION AT OR UNDER 20] </inputs> <task> Build the dashboard with: a header for [PILLAR TOPIC]; KPI cards (planned subtopics, covered, missing, median position of covered); a cluster map or card grid (subtopic, url or missing, position, clicks); a coverage bar; and a next-brief list of missing subtopics with a sample target query. Compute covered from my rule. Missing cards are visually loud. Clicking a gap card opens the brief line. Do not invent a full outline. One target query per missing subtopic, marked as a sample I can edit. </task> <constraints> - One self-contained HTML file; one CDN charting library only. - Cluster rows in labeled JS constants at the top so I can swap a cluster map and query export later. - Do not invent search volume I did not give. Mark missing volume as [VERIFY]. - Accessible contrast. </constraints> <format> Return the full HTML so I can click the working screen in the chat, then write the covered rule and which array to replace. </format>

You are a content SEO analyst who puts shipped posts next to the organic clicks they added in a fixed window. <context> I need an output-versus-organic-impact scorecard as one self-contained HTML file with inline CSS, vanilla JS, and a CDN chart library, seeded with a realistic ship log so I can click the working screen in the chat. </context> <inputs> - Calendar or site: [EDITORIAL CALENDAR] - Window after publish: [E.G. FIRST 28 DAYS] - Shipped posts: [10-16 TITLES WITH DATES OR "INVENT"] - Clicks in that window: [OR "INVENT A MIX OF LIFT AND ZERO"] - Zero-lift rule: [E.G. UNDER 20 CLICKS IN THE WINDOW AND AGE OVER THE WINDOW] </inputs> <task> Build the dashboard with: a header for [EDITORIAL CALENDAR]; KPI cards (ships, median clicks in window, zero-lift count, best ship); a bar of clicks in window by post; a weekly ships versus total new-post clicks line; a table (title, published, window clicks, current position, status ok/too new/zero lift); and a zero-lift list. Posts younger than the window are too new, not zero lift. Compute status in JS. Do not claim a post caused a site-wide click move. Per-URL window clicks only. Clicking a zero-lift row highlights the bar. </task> <constraints> - One self-contained HTML file; one CDN charting library only. - Ship rows in labeled JS constants at the top so I can swap the calendar and a Search Console pages export later. - Too-new labeled too new. Accessible contrast. - No causal site-wide language. </constraints> <format> Return the full HTML so I can click the working screen in the chat, then write window and zero-lift rules and which array to replace. </format>

You are a technical SEO analyst who builds Core Web Vitals boards from field metrics a crawl review can trust. <context> I need a Core Web Vitals dashboard as one self-contained HTML file with inline CSS, vanilla JS, and a CDN chart library, seeded with realistic LCP, INP, and CLS rows so I can click the working screen in the chat. </context> <inputs> - Site: [SITE] - URL groups or pages: [8-16 OR "INVENT REALISTIC PATHS"] - Metrics per row: [LCP, INP, CLS, GOOD/NEEDS IMPROVEMENT/POOR] - Thresholds: [LCP 2.5S, INP 200MS, CLS 0.1, OR "USE EDITABLE CONSTANTS"] - Period: [E.G. LAST 28 DAYS] </inputs> <task> Build the dashboard with: a header for [SITE]; KPI cards (percent good LCP, INP, CLS, failing URL groups); three vital cards with the threshold and a pass/fail; a table (url or group, LCP, INP, CLS, status per vital, overall); a failing-URL list where any vital is poor; and a short trend of percent good if you seed 6 weeks. Compute status from editable thresholds in JS. Clicking a failing URL highlights the row. Do not invent lab-only scores. Field metrics only. Show the thresholds on the page. </task> <constraints> - One self-contained HTML file; one CDN charting library only. - Vitals rows and thresholds in labeled JS constants at the top so I can swap a CrUX or Search Console vitals export later. - Poor rows flagged. Accessible contrast, not color-only. - Do not invent a live CrUX API. Sample data only. </constraints> <format> Return the full HTML so I can click the working screen in the chat, then write the threshold rules and which constants to replace. </format>

You are a technical SEO analyst who builds crawl budget watches that name wasted hits, not a vague crawl volume. <context> I need a crawl budget watch as one self-contained HTML file with inline CSS, vanilla JS, and a CDN chart library, seeded with realistic crawl rows so I can click the working screen in the chat. </context> <inputs> - Property: [PROPERTY] - Days: [E.G. LAST 14] - Daily crawled count: [OR "INVENT PLAUSIBLE"] - Sample crawled URLs with type: [MONEY, PARAM, FACET, 404, OTHER] - Waste types: [PARAM, FACET, 404] </inputs> <task> Build the dashboard with: a header for [PROPERTY]; KPI cards (URLs crawled, wasted hits, waste share, money-URL share); a daily line of crawl volume; a stacked bar of type mix; a waste list (url, type, hits, last crawled); and a table of types with hits and share. Compute waste as hits whose type is in my waste list. Clicking a type filters the waste list. Do not invent server capacity. Counts only. </task> <constraints> - One self-contained HTML file; one CDN charting library only. - Daily totals and URL-type rows in labeled JS constants at the top so I can swap crawl stats later. - Show the waste types on the page. Accessible contrast. - Do not invent a live log API. Sample data only. </constraints> <format> Return the full HTML so I can click the working screen in the chat, then write waste math and which arrays to replace. </format>

You are a technical SEO analyst who builds redirect chain boards that show hop count and the final status. <context> I need a redirect chain dashboard as one self-contained HTML file with inline CSS, vanilla JS, and a CDN chart library, seeded with realistic redirect paths so I can click the working screen in the chat. </context> <inputs> - Site: [SITE] - Sample start URLs: [10-16 OR "INVENT REALISTIC CHAINS"] - Hops and final status per start: [OR "INVENT 1-4 HOP MIX"] - Long-chain rule: [E.G. 3 OR MORE HOPS] - Loop examples if any: [OR "INCLUDE ONE LOOP"] </inputs> <task> Build the dashboard with: a header for [SITE]; KPI cards (start URLs, median hops, long chains, loops); a hop-count histogram; a chain table (start, hop count, final url, final status, loop yes/no); a selected-chain stepper that lists each hop; and a long-chain list using my rule. Compute hop count from a hops array on each row. A loop is when a URL repeats in the path. Clicking a long chain opens the stepper. Do not invent link equity math. Hops and status only. </task> <constraints> - One self-contained HTML file; one CDN charting library only. - Chain rows in labeled JS constants at the top so I can swap a redirect crawl later. - Loops flagged. Accessible contrast. - Do not invent a live crawl API. Sample data only. </constraints> <format> Return the full HTML so I can click the working screen in the chat, then write hop and loop rules and which array to replace. </format>

You are a technical SEO analyst who builds structured-data error boards from Search Console enhancement style counts. <context> I need a structured-data error dashboard as one self-contained HTML file with inline CSS, vanilla JS, and a CDN chart library, seeded with realistic rich-result rows so I can click the working screen in the chat. </context> <inputs> - Site: [SITE] - Types: [E.G. ARTICLE, PRODUCT, FAQ, BREADCRUMB, ORGANIZATION] - Valid, warning, and error counts per type: [OR "INVENT PLAUSIBLE"] - Sample error URLs: [OR "INVENT"] - Priority types I care about: [LIST] </inputs> <task> Build the dashboard with: a header for [SITE]; KPI cards (types tracked, error items, warning items, valid items); a stacked bar of valid / warning / error by type; a type table with those counts and error share; chips that filter to one type; and a fix list of sample URLs (url, type, issue, severity). Compute error share as errors / (valid + warning + error). Priority types I named sort first. Clicking a type filters the fix list. Do not invent rich-result impressions I did not seed. </task> <constraints> - One self-contained HTML file; one CDN charting library only. - Type counts and error URLs in labeled JS constants at the top so I can swap an enhancement report later. - Error share computed. Accessible contrast. - Do not invent a live Search Console API. Sample data only. </constraints> <format> Return the full HTML so I can click the working screen in the chat, then write error-share math and which arrays to replace. </format>

You are a technical SEO analyst who builds index hygiene boards from robots, noindex, and sitemap flags. <context> I need an index hygiene dashboard as one self-contained HTML file with inline CSS, vanilla JS, and a CDN chart library, seeded with realistic crawl flags so I can click the working screen in the chat. </context> <inputs> - Site: [SITE] - Sample URLs with flags: [robots allow/disallow, meta noindex yes/no, in sitemap yes/no, OR "INVENT A MIX"] - Mixed-signal rule: [E.G. IN SITEMAP AND NOINDEX, OR DISALLOW AND INDEX] - Money URL tags if any: [OR "NONE"] </inputs> <task> Build the dashboard with: a header for [SITE]; KPI cards (URLs scored, noindex, robots blocked, mixed signals); a bar of flag combinations; a sortable table (url, robots, noindex, sitemap, mixed yes/no, money yes/no); chips for mixed / noindex / blocked; and a mixed-signal list using my rule with a suggested check (remove from sitemap, drop noindex, or keep blocked). Compute mixed in JS from the flags. Clicking a mixed row opens the suggested check. Do not invent index counts from Search Console unless I seed them. </task> <constraints> - One self-contained HTML file; one CDN charting library only. - Flag rows in labeled JS constants at the top so I can swap a robots and noindex crawl later. - Show the mixed-signal rule on the page. Accessible contrast. - Suggested checks are rules, not a mandate. </constraints> <format> Return the full HTML so I can click the working screen in the chat, then write the mixed-signal rule and which array to replace. </format>