Changelog
What we've been building, month by month, on the road to launch.
Team accounts: multi-user roles (Viewer / Editor / Admin / Owner) with per-brand access control
White-label: per-brand logo upload, shown on shared reports and guest links (PRO+)
Sources rebuilt on real web-grounded AI citations and smart source discovery (replacing the mock data)
Comparison pages: honest, sourced side-by-side vs other AI-visibility tools, with a verdict card
Onboarding polish: per-page guided tours and a first-run empty-state hero on the dashboard
New 3-pane MDX documentation (in progress)
Prompt management: manual prompt selection at launch, add prompts (manual / bulk / AI-review), pin toggle
Persona-driven prompt generation with a dedicated writer model; avatar presets opened to the FREE plan
Richer transactional email templates; usage tracking and referral wiring
Model registry refresh; dashboard, KPI and drag-and-drop fixes
Audience avatars: persona profiles injected into LLM prompts (presets + custom, max 5 per brand)
Public API v1 (Agency): programmatic access to brands, topics, campaigns, visibility and avatars, with API keys and OpenAPI 3.1 spec
AI content engine: generate from a visibility gap or standalone, 8 formats, publish straight to CMS integrations
Redesigns: interactive homepage hero, full onboarding flow, sidebar grouped into Analyse / Présence / Actions with a usage bar
Stripe: reset credits on subscription renewal (invoice.paid); block account deletion while a paid subscription is active
Contact and roadmap pages, Google Search Console verification
First build: AI visibility tracking across ChatGPT, Claude, Gemini and Perplexity via a BullMQ async pipeline
Dashboard: KPIs, share of voice, timeline, competitor leaderboard, prompts and first opportunities view
Credits system: slot-based campaign cost, monthly quota per plan, daily cron
Supabase auth and onboarding (brand + topic + model selection); marketing skeleton and FR/EN i18n with feature gating
OpenRouter integration: ~57 models behind a single key via the OpenAI SDK, mock fallback when no key; Railway deploy
Suggestions? A missing feature?
Try it for free