audit_hint_category_list
ChatGPTUse this when you need the hint categories available for a finished audit before calling audit_hint_list. Returns data[] of category (snake_case), label, and hintCount. Do not use for listing individual hints or inventing an auditId. Requires auditId from project_audit_list on a finished audit.
audit_hint_list
ChatGPTUse this when the user wants SEO/technical hints (issues) in one category of a finished audit. Returns a paged list of hints with title, plain description, severity, coverage, URL counts/deltas, and snake_case urlFilterId for audit_url_list. The envelope carries hasPrevious and previousAuditId: per-hint previous counts are real comparisons only when hasPrevious is true. Resolve category first with audit_hint_category_list. Defaults to status=triggered (affecting ≥1 URL). Do not use for insights (use audit_insight_list), starting crawls, or inventing auditId/category. Requires auditId from project_audit_list on a finished audit and a valid category.
audit_insight_category_list
ChatGPTUse this when you need insight categories for a finished audit before calling audit_insight_list. Returns data[] of category (snake_case), label, and hintCount. Do not use for listing individual insights or inventing an auditId. Requires auditId from project_audit_list on a finished audit.
audit_insight_list
ChatGPTUse this when the user wants insights (informational findings) in one category of a finished audit. Returns a paged list sharing the hint shape (title, plain description, severity, counts, urlFilterId for audit_url_list). The envelope carries hasPrevious and previousAuditId: per-row previous counts are real comparisons only when hasPrevious is true. Resolve category first with audit_insight_category_list. Do not use for issue-style hints (use audit_hint_list), starting crawls, or inventing auditId/category. Requires auditId from project_audit_list on a finished audit and a valid category.
audit_lostfound_hint_list
ChatGPTUse this when the user wants how hints changed between this finished audit and the previous one. Returns paged per-hint deltas (new/added/removed/missing/remained counts, netChange, urlFilterId, title, severity) plus hasPreviousAudit and previousAuditId (the compared-against audit). If hasPreviousAudit is false, say this is the first audit with nothing to compare. Do not use for starting crawls or inventing an auditId. Requires auditId from project_audit_list on a finished audit.
audit_lostfound_url_list
ChatGPTUse this when the user wants sample URLs that are new, missing, or remained versus the previous audit. Returns up to 50 sample rows (url, urlHash, state) plus hasPreviousAudit and previousAuditId (the compared-against audit); not a full export. Optional state: remained, new, or missing. Optional urlFilterId from audit_lostfound_hint_list. Do not use for inventing auditId or treating the sample as complete. Requires auditId from project_audit_list on a finished audit.
audit_report_get
ChatGPTUse this when the user wants one report by slug (e.g. onpage) for a finished audit. Returns score, hints (id, hintId, urlFilterId, sampleUrls), relatedReports.reports[] with reportSlug, and data[] tables — render tables rather than sampling them. For URL drilldown pass hints[].urlFilterId to audit_url_list; do not use id or hintId for URL lists. Do not use for starting crawls or inventing auditId/slug; resolve slug via audit_report_list.data[].slug or relatedReports.reports[].reportSlug. Requires auditId from project_audit_list.data[].auditId on a finished audit and a known slug.
audit_report_list
ChatGPTUse this when the user needs which report slugs are available for a finished audit. Returns data[] of slug, title, and description for fetchable reports only. Pass slug to audit_report_get. Prefer audit_report_summary or audit_report_seo for common overviews. Do not use for starting crawls or inventing auditId. Requires auditId from project_audit_list.data[].auditId on a finished audit.
audit_report_seo
ChatGPTUse this when the user wants the site-wide SEO report for a finished audit. Returns SEO score and report tables, plus relatedReports.reports[] with reportSlug for audit_report_get. Do not use for starting crawls or inventing auditId. Requires auditId from project_audit_list.data[].auditId on a finished audit.
audit_report_summary
ChatGPTUse this when the user wants the site-wide audit overview (crawl summary) for a finished audit. Returns the crawl report snapshot: scores, totals, breakdowns, audit.auditId, and audit.previousAuditId when present. Do not use for starting crawls, other report slugs (use audit_report_get), or inventing auditId. Requires auditId from project_audit_list.data[].auditId on a finished audit.
audit_url_get
ChatGPTUse this when the user wants full column data for one URL in a finished audit. Returns data as a dictionary of PascalCase columns (Url, HttpStatus, WordCount, …); null columns omitted. Prefer urlHash from audit_lostfound_url_list when available; otherwise pass url. Do not use for listing many URLs (use audit_url_list sample) or inventing auditId. Requires auditId from project_audit_list on a finished audit, plus url or urlHash.
audit_url_list
ChatGPTUse this when the user wants sample URLs affected by a specific hint or insight in a finished audit. Returns up to 50 sample rows of { Url } only (not paginated, not the full list). Take urlFilterId from audit_hint_list, audit_insight_list, audit_lostfound_hint_list, or audit_report_get hints[].urlFilterId. Do not use for the complete URL export, starting crawls, or guessing auditId/urlFilterId. Requires auditId from project_audit_list on a finished audit, and urlFilterId as above.
project_audit_list
ChatGPTUse this when the user needs audits for a known project, or an auditId for finished-crawl tools. Returns a paged data[] of audits (auditId preferred, id legacy alias, projectId, status, urlsFound, urlsAudited, completedAt, durationSeconds) plus pagination, newest first. Filter status=finished to find audits ready for hints, reports, and URLs. Pass auditId to audit_* tools. Do not use for starting, pausing, stopping, or deleting audits, or inventing an auditId. Requires projectId from project_list.data[].projectId.
project_history_list
ChatGPTUse this when the user wants historical metric trends across completed audits for a project. Returns a paged list of snapshot objects with snake_case metric keys (site_score, internal_links, crawl_depth, start_date, …). Each snapshot carries audit_id, has_previous, and previous_audit_id for drilling into either audit with the audit_* tools; has_previous false means it is the earliest snapshot. Present trends across snapshots rather than only the latest value. Do not use for live crawl control or inventing a projectId. Requires projectId from project_list.data[].projectId.
project_list
ChatGPTUse this when the user wants to find Sitebulb projects by name or URL, or needs a projectId for later tools. Returns a paged list (data[] + pagination) with projectId (preferred), id (legacy alias), name, url, auditCount, and lastCrawled. Pass projectId from a matching row to project_audit_list or project_history_list. Do not use for starting crawls, changing settings, deleting projects, or general SEO advice without Sitebulb data. If no project name or URL was given and a list is not what the user wants, ask which project they mean. Empty data[] with totalItems 0 means no projects match the search (or the account has none) — say so; do not invent a projectId. Empty data[] on a later page with totalItems > 0 means you are past the last page — request a lower page, do not treat it as "no projects".