Every part of the
workflow, forged.
Seven categories. One coherent pipeline. From the moment you drop in a URL to the moment a writer downloads a brief, every step is deliberate.
The depth differentiator
A multi-phase background workflow that turns a niche into a prioritised opportunity board. Every step is observable, retriable, and grounded in real SERP data — never a synthesised guess.

AI niche explication
A URL or a sentence becomes a structured market profile — target customer, core problem, related terms, language, country. Edit any field. Refine in plain English.
Async multi-phase workflow
One button kicks off a durable background run. Close the tab, come back tomorrow — your run resumes where it left off.
Keyword discovery from your niche
We expand niche-derived seeds through autocomplete, fan out hundreds of candidates, then score every one for relevance and intent.
Live SERP data, not stale tables
We collect the top results per keyword in real time, batched in parallel. Your scoring reflects what the market looks like today.
AI relevance and intent scoring
Each keyword carries a 1–5 niche-fit score, an intent classification, and a volume bucket. You stop reading raw numbers and start reading meaning.
Coverage-aware dual-pool clustering
Keywords are labelled covered or uncovered against your real content, then clustered in two separate pools. Gap clusters surface what to write next; covered clusters surface what to refresh.
Editorial cluster naming
AI names every cluster as a working headline, not a keyword string. Your roadmap reads like an editorial calendar.
Real-time progress stream
A live, animated UI follows the workflow: keyword discovery, SERP collection, competitor profiling, clustering. Network drops? The stream reconnects.
Built for the way you actually work
Switch projects without losing your place. Every page is anchored to your active project. Your workspace remembers what you were doing and where you were doing it.

Create a project from a URL or a sentence
Drop in your homepage URL — or describe your niche in plain English. Either way, your project ID lands near-instantly and the workflow starts in the background.
Free sitemap mapping on every project
We pull robots.txt, the sitemap index, every gzipped child sitemap, and read up to a hundred of your pages. We tell you what you already cover, before you spend a credit.
Niche confirmation gate
No paid run starts until you confirm we read your business right. The CTA stays disabled with an honest tooltip — never a silent failure.
Active-project workspace
Switch projects on Opportunities and you stay on Opportunities. Cross-tab sync keeps every window in step. No context loss, ever.
On-demand asset re-mapping
Published five new articles? Hit re-map. We re-read your sitemap, re-extract target keywords, and feed them back into your next run. Always free.
Default ignored competitors
YouTube, Reddit, Pinterest, Facebook and a dozen other noise domains are pre-flagged as ignored. You curate signal, not chaff.
See what to work on. In priority order.
A composite opportunity score blends volume, difficulty, intent, and niche-fit — boosted when your important competitors are already ranking. The board is tiered so the right cluster lands at the top of the page.

Composite opportunity scoring
Volume bucket × difficulty × intent × niche-relevance. One score, one ranking, one prioritised list.
Important-competitor boost
Flag a competitor as important and clusters where they rank get a 1.25× score boost. You go where the fight already is.
Coverage badges
Covered, Partial, or Gap — labelled live from SERP signals plus content-match against your sitemap. Stop drafting articles you have already shipped.
Goal filters
Switch the board between Fast Traffic Wins and Revenue-Driving Content. The same data, framed by what you are actually trying to do.
Cluster detail page
Every cluster opens to its keywords, the full SERP, the top three competitors on a podium, and any of your own articles that overlap.
Editorial scoring on covered content
Healthy. Needs refresh. Needs rewrite. Should prune. No coverage. Every existing page gets an editorial verdict — no spreadsheet required.
Per-keyword SERP detail
Drill into any keyword to see the live top-10, related keywords, and the cluster it belongs to. Stop tab-switching to Google.
A Kanban that ships
Five stages. Writer-ready briefs on every card. Download triggers a status change. Your roadmap is auditable, defensible, and always one click from execution.

Five-status Kanban
Planned → Generating Brief → Brief Ready → Out for Writing → Monitoring. Action-driven transitions only — you cannot mis-set a status by drag-and-drop.
Auto-create plan from top opportunities
One click turns your top thirty opportunities into a populated board. Edit, add, reorder before you commit.
Writer-ready briefs (v2)
H2 and H3 outline, primary and secondary keywords, intent alignment, FAQ questions, reference articles from real SERP data, acceptance criteria. Not an LLM hallucination — a structured deliverable.
Internal links classified
Briefs separate links you already have ([Existing]) from links worth building ([Desired]). Your editorial calendar feeds your link strategy automatically.
Two download formats
PDF for your human writers. Markdown optimised for AI agents. The same brief, packaged for the consumer.
Manual articles welcome
Add a freeform suggestion that did not come from a cluster. The pipeline does not care where the idea started — it cares whether it ships.
Know who you are fighting
AI-profiled companies, ranked pages, and per-niche edge analysis. Curate once — your important and ignored flags survive every re-run.

AI competitor profiling
Each competitor is read by an AI agent — company type, offering, pricing model, sales motion, strengths, weaknesses, differentiators. A dossier, not a domain.
Per-niche edge analysis
For each cluster: direct, indirect, or substitute competitor; relevance 0–100; one-line insight. You stop guessing whether they actually compete.
Importance management
Mark a domain Important and it boosts your scoring. Mark another Ignored and it disappears from your board, podium, and brief inputs.
Curation survives re-runs
Your importance flags are sticky by design. Run research again next month and your judgment is preserved, not overwritten.
Ranked pages explorer
Filter every page that ranks anywhere in your data — by domain, by ownership, by cluster. Your competitor's playbook, in one table.
Page dossier
Open any ranking page to see its rankings, SERP positions, target keyword analysis, and heading outline. Reverse-engineer the article without guessing.
A portfolio cockpit
Every project rolls up into a cross-project home. Pipeline velocity. Cross-portfolio Next Steps. Outliers flagged in gold. The first thing you see is the first thing you should act on.

Outcome metrics, not vanity numbers
Briefs ready, with weekly delta. Untapped opportunity count. Live keyword footprint. Available credits. The four numbers a CEO actually wants to see.
12-week pipeline velocity
A stacked-area chart of work moving through your Kanban over the last quarter. Volume and distribution, on one slide.
Top-five Next Steps
Cross-project, severity-ranked, gold-dot urgent. A 5-minute morning scan tells you what to do today across every project you run.
Portfolio table with mini-bars
Mini-bars for opportunity score and pipeline. Staleness colour: amber after 30 days, red after 60. Triage at a glance.
Per-project overview
Stat cards, top-five priority articles, top-five fiercest competitors, contextual next-step guidance. The same surface, scoped to one client.
Pinning system
Pin queries, pages, clusters, or companies. Reorder them. Bulk-promote pinned items into your content plan. Your judgment, made first-class.
Drive the product from your AI assistant
VarynForge ships a Model Context Protocol server. Claude Desktop, Cursor, your own scripts — every AI agent you trust can list projects, generate briefs, and update Kanban without opening the dashboard.
Procedure-level opt-in
Every tRPC procedure decides whether to expose itself to AI agents. The MCP surface is curated, not auto-generated.
OAuth, dynamic client registration
Standards-compliant authentication: RFC 8414 discovery, browser-based authorisation, token exchange. Your assistant connects in seconds.
Thirteen agent-ready tools
List projects, get a project, get an article suggestion, generate an article brief, update an article status — and more. The pipeline is scriptable end-to-end.
OAuth discovery proxy
A dedicated /.well-known/oauth-authorization-server endpoint so any MCP-compliant agent finds the right metadata on the first try.
What we do not do
(yet, or ever).
We will not pretend. Here is what other SEO tools ship that we do not.
No backlink graph
We do not crawl the web for backlinks. No domain authority, no link gap. If you need that, Ahrefs and Moz exist.
No daily rank tracking
Positions are snapshots from your last research run, not a daily poll. Monitoring is a Kanban label here, not a tracker.
No on-page editor
Briefs are downloaded and consumed in your CMS. We are not Surfer's content editor.
No site audit
No Core Web Vitals, no broken-link check, no canonical audit. We are an SEO research tool, not a technical-SEO platform.
No multi-locale (yet)
The pipeline is English-first. Other languages are on the roadmap, not in production.
No keyword-database import
Search volume comes from live SERP signals and AI inference, not a backlink-indexed database. Numbers will not match Ahrefs to the unit.
Drop in a URL.
Watch the workflow run.
Project creation, niche analysis, and sitemap mapping are always free. See the system understand your business before you spend a credit.