Product-led SEO: research, planning, monitoring and agent tools
What changed
- Plan from a product URL without Search Console data: an editable product profile, autocomplete topic discovery, a guard that flags topics overlapping content you already have, and Fill my month, which schedules up to 12 topics without drafting them.
- Research: a question tree built from autocomplete, a SERP brief and term comparison (needs an operator Brave Search API key), a competitor content gap from a public sitemap, and optional DataForSEO volume and ad competition.
- Plan structure: a topical map of pillars and spokes computed from the plan, alternatives, use-case and integration templates, dataset building from the profile, and a weekly topic check.
- Measure: a rank tracker for up to 25 keywords (Brave results), a technical audit of up to 25 sitemap pages, and an AI search visibility check that spends your OpenRouter credits.
- Post editor: a Quality check, Compare with top results, and no-AI Repurpose; llms.txt generation and BlogPosting and FAQPage JSON-LD, with FAQ markup added to Ghost drafts automatically.
- MCP: four read-only research tools (analyze_post, discover_topics, research_keyword, find_content_gaps) and a documented agent playbook that stops at a CMS draft.
Not verified against live providers: Brave, DataForSEO, OpenRouter web search and OpenAI embeddings were exercised with stubbed responses in testing. Draft-only delivery is unchanged.