https://api.rankraft.ai/v1. Every request sends the key in X-API-Key. Reports are POST with a JSON body and return { info, data }; discovery endpoints are GET and return { data }. Full parameters are in the REST API tab; the rules that apply to all of them are in Conventions & gotchas.
Discovery
Answer reports
All takesite_id, start_date, end_date (inclusive, UTC, at most 92 days apart), an optional filter and cursor, and a limit of up to 100.
POST /v1/reports/visibility
Visibility Score and its parts, for your brand or for you and your competitors.
Recipes: score with change, headline and daily, over time, leaderboard, compare competitors, segment.
POST /v1/reports/citations
Which domains or pages the engines cite, and who owns them.
Every row also carries
domain, owner (own, competitor, third_party), source_kind and rank.
answers counts distinct answers citing the domain; citations counts every link, so a page linked twice in one answer is 1 answer and 2 citations. rank follows answers. citation_share is citations ÷ all citations in the same group after the filter.POST /v1/reports/sentiment
The tone of answers that name you.
POST /v1/answers
The stored answers themselves, newest first: text, brands named, citations and sentiment. Narrow with platforms and prompt_ids, trim with include. At most 50 a page.
AI traffic reports
Both takesite_id, start_date, end_date (at most 400 days apart), an optional timezone (IANA, default UTC), filter, sort, and a limit of up to 500.
POST /v1/reports/bots
AI and search crawler hits.
Spoofed hits (
verified = failed) are left out unless your filter mentions verified.
POST /v1/reports/referrals
People arriving from AI assistants.
Recipe: AI crawler traffic to a warehouse.
Prompts
Need a key with Read and manage prompts access. Prompts are never deleted over the API; switch them off instead.
Recipe: Manage prompts from a CSV.