featurebeta
GEO Agent
Multi-agent fleet that runs autonomously: Monitor, Research, Content, and Coordinator agents scan prompts, detect gaps, generate fixes, and wait for your approval before applying changes. L0–L3 automation levels.
Changelog
Shipped continuously. Every feature listed here is live in the product.
Multi-agent fleet that runs autonomously: Monitor, Research, Content, and Coordinator agents scan prompts, detect gaps, generate fixes, and wait for your approval before applying changes. L0–L3 automation levels.
Rebuilt the prompt engine across all 7 supported AI models. Improved citation detection accuracy, better source attribution, and more reliable gap classification. Resolved 18 tracked issues.
External API with scoped bearer tokens, cursor-based pagination, and per-key rate limiting. Covers visibility, prompts, scans, and optimize endpoints. Agent approval actions excluded by design.
Role-based access control across all settings and feature pages. Custom role dropdown with inline permission descriptions, seat limits per plan, and bulk invite flow with audit log.
Test content changes before publishing. Simulate how ChatGPT and Claude would respond to your tracked prompts after proposed edits — without waiting for a live scan.
AI-generated content briefs based on what gets cited in your category. Pulls live citation patterns from scans and produces structured briefs mapped to specific prompt gaps.
Browsable index of 36+ GEO prompts across categories and intent types. Daily-updated trending queries, filter by engine and intent, and track which prompts drive citations in your space.
Track how AI shopping assistants rank and recommend your products. Separate from brand visibility — covers product-level citations, comparison queries, and purchase-intent prompts.
Compare citation share against up to 5 competitors per prompt. Shadow competitor detection using 2× presence × 2× recency × 2× engine coverage threshold.
Visibility scoring across ChatGPT, Perplexity, Claude, Gemini, and more. Per-engine breakdowns, trend graphs, and source attribution showing exactly where citations come from.
Real-time notifications for visibility drops, new competitor detections, and scan completions. Configurable thresholds per engine with workspace-scoped webhooks.