create_cluster
ChatGPTCreate a new topic cluster and enqueue AI-assisted translation. Requires a name. Locale defaults to the tenant's default locale.
Overview
Flinty helps teams inspect SEO content and search performance, monitor AI visibility and citations, create topic clusters, generate or refresh articles, schedule publication, and audit public pages through ChatGPT.
Tools
enrich_reddit_thread
ChatGPTFetch and store the full content of a Reddit thread (title, body, score, comments count). Cached 24h.
get_article
ChatGPTGet full content and metadata of an article by ID.
get_cluster
ChatGPTGet details of a topic cluster including all its subjects and article statuses. satellites_published/pillar_published only cover this cluster, not the tenant's other content.
get_geo_analysis
ChatGPTGet GEO (Generative Engine Optimization) analyses — brand mentions across ChatGPT, Perplexity, Gemini, Grok and Claude. Per-engine results are returned. Gemini and Grok answer from model knowledge and carry no citation URLs.
get_metrics
ChatGPTGet Google Search Console metrics summary (clicks, impressions, top queries). May query Google live and refresh the stored OAuth connection state when required.
get_top_pages
ChatGPTGoogle Search Console breakdown by page (clicks, impressions, average position) for a window, each row resolved to its published article when the URL is a blog article. Queries Google live for the exact window (not the cumulative snapshot cache). Default window is the last 30 days ending yesterday; pass period ('7d'/'30d'/'90d') or explicit start_date/end_date (YYYY-MM-DD).
list_articles
ChatGPTList articles. Optionally filter by status (draft, ready, scheduled, published, archived).
list_backlink_opportunities
ChatGPTIdentify domains that cite your competitors in Perplexity responses but not you — high-value targets for getting mentions.
list_clusters
ChatGPTList all topic clusters for this blog with their status and metrics. satellites_published/pillar_published only cover this cluster's own subjects — they don't include articles published outside the cluster pipeline (YouTube import, web import, Reddit, etc). For a tenant's true total published article count, use list_articles. Response is capped (default 50, max 200) — check has_more before assuming the list is complete.
list_competitors
ChatGPTList tracked competitors with their brand names, URLs, and total GEO citation counts.
list_llm_sources
ChatGPTList domains most cited by ChatGPT, Perplexity and Claude (the engines returning citation URLs) when answering your GEO prompts, ranked by citation frequency. Gemini and Grok are excluded because they return no sources.
list_missed_citations
ChatGPTFind GEO analyses where your domain was cited by ChatGPT, Perplexity or Claude (the engines returning citation URLs) but your brand name was NOT mentioned in the response — missed brand attributions.
list_page_audits
ChatGPTList SEO page audits previously run on this blog.
list_reddit_threads
ChatGPTList Reddit threads cited by Perplexity in your GEO analyses and store missing thread records discovered from those citations.
list_sites
ChatGPTList all sites (tenants) in your organization. For agency accounts managing multiple sites.
publish_article
ChatGPTSchedule a ready article for automatic public publication. Once due, it may also distribute a post to connected social accounts. Optionally provide a scheduled_at datetime.
refresh_article
ChatGPTTrigger a rewrite of a published article in the tenant's default locale. The refreshed content replaces the public version after a recoverable snapshot is stored.
run_page_audit
ChatGPTRun an AI-generated SEO audit on a URL: fetches the page's title and main content, then an LLM analyzes title, meta description, headings, content quality, and internal linking to produce a score (0-100) and actionable recommendations. This is a qualitative content assessment — not a technical/performance audit (no Core Web Vitals, load time, or Lighthouse-style metrics). Cached 24h.
trigger_generation
ChatGPTLaunch AI-assisted article generation for a cluster. The cluster must be in 'draft' or 'planned' status. Generation runs asynchronously through external AI services.
Capabilities
Writes
Example Prompts
Click any prompt to copy it.
