acknowledge-hosted-self-check
ChatGPTRecord that a person has checked a hosted page that the automated fact check never read. It stores a human attestation against the credential owner that someone has actually read the page.
add-competitor
ChatGPTAdd competitor domains to tracking. A domain already tracked keeps its display name and aliases and is listed under already_tracked.
archive-article
ChatGPTArchive an article. It leaves the article lists and its slug is replaced with archived-<id>, which frees the old slug for reuse. The content is kept.
archive-prompt
ChatGPTArchive a prompt. It leaves future visibility runs and stops counting toward the prompt quota; its past scores and answers are kept, and restore-prompt brings it back. It keeps its topic.
compare-visibility-runs
ChatGPTCompare 2-4 visibility runs across 1-20 slices in one table, at most 40 cells (runs × slices), each cell scored exactly as the Visibility page scores that run. A run is a run id or a YYYY-MM-DD day (that day's latest completed run, in time_zone). A slice is a prompt query type (branded = Branded, non_branded = Non-branded, competitor_branded = Competitor-branded; default non_branded), optionally narrowed by topic_id, custom filters (keys and value slugs from list-prompt-filters) and engine, with an optional row label. Each cell gives AI AuraScore (0-100), Mention Rate and Share of Voice (percentages, 0-100), Rank and Sentiment for the tracked brand. include_competitors adds competitor rows for the first slice, the top 30 by the newest run's Rank, with competitorsOmitted counting the rest; competitorSlice names the slice the competitor rows belong to. Runs come back oldest first. scopeTooSmall means the slice has too few prompts for an AI AuraScore or Rank; noAnswers means the run has no answers in that slice. A run's notes name scoring rules it predates; noteText explains them. An unknown slice key or a topic_id not on this domain is rejected. The result has one row per slice and one column group per run, labelled by its day. More than 4 runs is rejected.
create-article
ChatGPTCreate an article on a domain from a finished Markdown body, with no AI generation. It counts toward the articles quota for the billing period and is refused once the quota is used up. The slug is made from the title unless given; a slug already used on the domain is rejected. Status is ready (default) or draft. Needs a user-tied credential. Returns the new article id, slug and status.
create-prompt
ChatGPTCreate a visibility prompt, the question PingAura asks AI engines on each run, for one domain. The topic is created if it does not exist. An active prompt joins the next visibility run and counts toward the prompt quota, so creation is refused at the quota; a draft prompt does neither. A prompt with the same text on the domain is refused.
create-prompt-topic
ChatGPTCreate an empty prompt topic for one domain. Topic names are unique per domain, ignoring case.
delete-competitor
ChatGPTStop tracking a competitor on its domain. Its past results stay in completed runs; future runs no longer score it, and the domain's prompts are relabelled for whether they name a competitor.
delete-prompt-topic
ChatGPTDelete a prompt topic. The topic must hold no prompts, archived ones included: an archived prompt keeps its topic.
dispose-hosted-claim
ChatGPTRecord a decision about one claim a fact check flagged on a hosted page. Every flagged claim must be decided before the page can be approved. Claim indexes come from get-hosted-page.
get-account-info
ChatGPTGet the connected workspace's name, slug and creation date, and its active subscription's status and billing period. Read only.
get-article
ChatGPTGet one article in full: title, slug, status, keywords, the Markdown body, meta description, summary, word count and where it was published. Read only.
get-audit-issues
ChatGPTGet the issues found in one site audit, highest impact first. Each issue has a severity, a category and the issue_fingerprint that update-issue-status takes.
get-domain
ChatGPTGet full details for a specific domain, including icon_url, brand_aliases and sitemap scan state. The pingaura://domains/{domainId} resource is a lighter read-only view of the same domain.
get-hosted-page
ChatGPTGet one hosted page: its body, the evidence behind it, and the claims a fact check flagged.
get-openai-ads-account
ChatGPTGet account-level metadata for the ChatGPT Ads account connected to a domain: id, name, currency, timezone, preview URL.
get-openai-ads-account-insights
ChatGPTGet account-level insights (spend, impressions, clicks, CTR, CPC, CPM) over a date range. aggregation_level returns one row per campaign, ad group or ad instead.
get-openai-ads-ad
ChatGPTGet a single ChatGPT Ads ad by ID, including creative.
get-openai-ads-ad-group
ChatGPTGet a single ChatGPT Ads ad group by ID.
get-openai-ads-ad-group-insights
ChatGPTGet insights scoped to a single ad group over a date range.
get-openai-ads-ad-insights
ChatGPTGet insights scoped to a single ad over a date range.
get-openai-ads-campaign
ChatGPTGet a single ChatGPT Ads campaign by ID, including budget, schedule, status, and targeting.
get-openai-ads-campaign-insights
ChatGPTGet insights scoped to a single campaign over a date range. aggregation_level returns one row per ad group or ad instead.
get-prompt-scores
ChatGPTGet visibility scores for tracked prompts/queries. Each row carries prompt_type: 'branded' names only our brand, 'competitor_branded' names a competitor (with or without us), 'non_branded' / 'unclassified' names neither and counts toward the non-branded headline. Results are paged: total is the full match count, and while has_more is true further pages come back for after set to next_cursor, so a single page is not the full set.
get-quota-usage
ChatGPTGet used, limit and remaining for each quota: prompts (active prompts), articles (created this billing period), site audits and chat messages. A null limit means unlimited.
get-site-health-summary
ChatGPTGet a computed site health aggregate for a domain: total issue counts and per-severity breakdown (critical/high/medium/low always present, zero when none) across the 5 most recent completed audits. The pingaura://domains/{domainId}/health resource holds the raw audit metadata.
get-visibility-score
ChatGPTGet the exact visibility score for the scope you name: prompt type, topic, custom filters and engine, in one run or a date window. Every figure is scoped to the arguments given and matches the dashboard for the same selection. Each field matches a dashboard name: score = AI AuraScore (0-100), coverage = Mention Rate (a 0-1 ratio, mentions / answers), share_of_voice = Share of Voice (a 0-1 ratio), mentions = answers that mention you, mention_count = mentions and citations behind Share of Voice; answers and prompts are the counts in scope. It is NOT equivalent to averaging get-prompt-scores: per-prompt scores do not average into the scoped score, and this tool returns the scoped figure directly. Scope is run_id when given, else the latest completed run of run_date, else the latest completed run for range latest_run, else the last range days (default 15d, the same default period as the dashboard). prompt_type defaults to non_branded (prompts naming neither us nor a competitor); branded names only us, competitor_branded names a competitor with or without us, and each prompt is in exactly one. Valid filters keys and value slugs are the ones list-prompt-filters returns; an undeclared key is rejected with the valid options. Score and share_of_voice are null when the scope has nothing to score. The response scope.run_id is the run scored (null for a date window). compare-visibility-runs compares runs or slices side by side.
get-visibility-summary
ChatGPTGet the visibility summary for a specific run, returning a single object with scores, coverage %, engine breakdown, and competitor comparison. The run_id comes from list-visibility-runs. geo_visibility_score and competitor_scores are the NON-BRANDED score (prompts naming a brand, ours or a competitor's, are excluded); the _branded fields hold the score for prompts naming only us, and the _competitor_branded fields hold the score for prompts naming a competitor, including ones that also name us. Both pairs are null when the run had none of that type. When competitor_branded_prompt_count is null the run predates the split: geo_visibility_score_branded, competitor_scores_branded, branded_prompt_count and each engine_scores entry's geo_branded / competitor_scores_branded are then null too, because the stored values mixed both brand types. An engine_scores entry whose geo is null answered only brand-type prompts: it has no non-branded data (not a score of 0), only its brand-type fields. total_answers, answers_with_presence, coverage and average_answer_score are non-branded too: they count only the prompts behind geo_visibility_score, not every prompt in the run. By default each competitor list holds the 10 highest-scoring brands plus the tracked brand, competitors_omitted counts the rest, and engine_scores entries carry scores without per-engine competitor lists; detail "full" returns every list.
list-articles
ChatGPTList the AI-optimised articles of one domain, newest first, optionally filtered by status (draft, generating, ready, published, failed). Each row has the title, slug, topic, keywords, word count, summary and publish date, but not the body. Read only.
list-competitors
ChatGPTList the competitor domains tracked for one domain, with each competitor's id, display name and aliases. The display name and aliases decide which AI answers count as a mention of that competitor. Read only.
list-domains
ChatGPTList the domains tracked in the connected PingAura workspace, with each domain's id, name, verification status, industry, target geography and sitemap scan state. Every other domain tool takes one of these ids. Read only.
list-hosted-pages
ChatGPTList the pages written for a domain and the review queue they sit in.
list-leads
ChatGPTList leads captured on a domain's lead pages, newest first. Pass since (ISO time) to get only leads created after it. Returns id, created_at, state, page, AI source, score (0-100) and company. Name, email, phone and message are included only for workspace owners on Enterprise Reporting; otherwise they are null and contact_details is 'enterprise_owners_only'. Requires the Lead Pages feature.
list-openai-ads-ad-groups
ChatGPTList ad groups belonging to a ChatGPT Ads campaign, with status and bidding config.
list-openai-ads-ads
ChatGPTList ads belonging to a ChatGPT Ads ad group, with creative, status, and review_status.
list-openai-ads-campaigns
ChatGPTList campaigns on the connected ChatGPT Ads account with status, schedule, budget, and targeting.
list-prompt-filters
ChatGPTList the custom filters a domain declares: each key, its human label, and its valid values as value slug plus label. get-prompt-scores filters attributes by these value slugs; the label is a separate human string. attributes comes back empty when the plan does not include custom filters or the domain has declared none.
list-prompt-topics
ChatGPTList the prompt topics of a domain with their ids, sorted by name.
list-site-audits
ChatGPTList site health audits for a domain, newest first. Each audit covers one page and gives its url, scores (technical, content, aeo), PingAura score, issue count, Lighthouse performance on mobile and desktop, and Core Web Vitals status.
list-site-pages
ChatGPTList the pages discovered for a domain, sorted by URL, with each page's tracked flag, issue count, latest scores (technical, content, aeo) and last audit time.
list-visibility-runs
ChatGPTList visibility analysis runs for a domain, newest first. Each run lists the engines it asked; engine_shortfall, when present, names the engines that answered fewer prompts than expected (dropped of expected). With one_per_day, list only completed runs, one per day (that day's latest, the run the Visibility page shows for it), each with its day; get-visibility-score and compare-visibility-runs accept those run ids and days. one_per_day reads at most 500 of the newest runs, so on a domain with hundreds of runs in a few days it may return fewer than limit days.
move-prompt-to-topic
ChatGPTMove a prompt to another topic of the same domain. Its text, scores and history stay as they are; topic-level scores reflect the change from then on.
publish-article
ChatGPTPublish an article. Auto-routes to a connected CMS integration when exactly one is connected for the article's domain. With several connected, target_provider picks one. With none, deployed_url records where the article was deployed by hand. In all cases the article's local status is set to "published" with a timestamp.
reject-hosted-page
ChatGPTReject a hosted page and close its topic, so discovery will not propose it again. A page that is already live starts retiring. request-hosted-page-regeneration keeps the topic and asks for a new version.
request-hosted-page-regeneration
ChatGPTSend a hosted page back to be written again, carrying a note on what to change. The topic stays queued. A page that is already live starts retiring, because a draft cannot be served.
restore-prompt
ChatGPTRestore an archived prompt to active, so it joins the next visibility run. Refused when the workspace is at its prompt quota.
retire-hosted-page
ChatGPTTake a live page off the customer's domain. Without a redirect the URL answers 410, which tells an answer engine it is gone rather than missing. With one it 301s to another page that is already live. Requires the account owner.
run-site-audit
ChatGPTQueue a site health audit of a domain's tracked pages, or of the page ids given. Each page audited uses one unit of the site audit quota, and the request is refused when it would exceed the quota. Audits run in the background, fetching each page; results appear in list-site-audits once complete.
set-prompt-filter-definitions
ChatGPTReplace the custom filters a domain declares. Send the complete intended list, not a patch: anything you leave out is removed. Removal is refused while prompts are still tagged with it, because nothing could then filter or label those values. The current list is what list-prompt-filters returns. At most 5 filters, 50 values each; keys are lowercase letters only and value slugs lowercase alphanumeric.
sync-sitemap
ChatGPTQueue a sitemap scan for a domain to discover and track pages. Required before running a site audit when no pages exist yet.
update-article
ChatGPTUpdate an article's title, slug, status (draft, ready or published), keywords, meta description or summary. Only the fields sent change; each one overwrites its old value. The body is not editable here.
update-competitor
ChatGPTUpdate a competitor display name or aliases. These names decide which answers count as a mention of the competitor; its domain name only counts when it matches the display name.
update-issue-status
ChatGPTMark a site health issue done or reopen it. The status applies to the issue on that domain, identified by its fingerprint, across audits.
update-prompt-filters
ChatGPTSet the custom filter values on an existing prompt. Merges: keys you omit keep their current value, and a key sent as "-" is cleared. The declared keys and their valid values are the ones list-prompt-filters returns; an undeclared key or value is refused rather than dropped. Prompt text is immutable; move-prompt-to-topic changes the topic and archive-prompt / restore-prompt change the status.
update-prompt-topic
ChatGPTRename a prompt topic. Its prompts and their scores are unchanged.