A working prototype of the full agentic SEO/GEO pipeline: research, briefing, writing, quality gating, publishing, citation monitoring and continuous optimization, with human oversight at the points that matter and a full audit trail per client.
Open the interactive demo → How it worksOptimize feeds newly discovered content gaps back to Research — the loop runs monthly, indefinitely. Below-threshold drafts bounce between Quality and Writer up to a configurable number of attempts, then escalate to a human. Every action lands in the per-client audit log.
Keyword expansion, conversational long-tail discovery, competitor gap analysis, AI-intent mapping and a hub-and-spoke topical map. Long-tail first is enforced: high-competition terms are excluded from month one. Output: a prioritized monthly content plan (8–12 articles).
One brief per planned article: target keywords with fan-out queries, word count, heading structure, internal link targets (spokes → pillar, pillar → spokes), schema type, competing URLs and a unique-angle directive.
Long-form, people-first drafts built for citation: a direct answer in the first 150 words, named entities, sourced facts, structured Q&A, a mandatory FAQ, JSON-LD schema and 2,000+ words. Rewrites target the exact criteria Quality failed.
Scores every draft 0–100 against the GEO rubric below. Below threshold → returned to the Writer with the failed criteria. Still failing after max attempts → escalated to a human with one-click resolution.
Verifies robots.txt allows GPTBot, ClaudeBot, PerplexityBot and Googlebot (blocked = escalate, publish nothing), pushes the post with metadata, canonical, Open Graph and schema, regenerates llms.txt, pings IndexNow and Search Console.
Weekly per-keyword rankings, GA4 traffic, AI-bot crawl activity and per-engine citation status (ChatGPT, Gemini, Perplexity, Google AI Overview). Citations are probabilistic, so every run re-checks; drops are flagged the moment they happen.
Rewrites intros, extends FAQs, reinforces internal links and refreshes facts on flagged pages, republishes them, and feeds new content gaps back to Research — closing the loop.
Structured JSON handoffs with PASS / FAIL / NEEDS_REVIEW / ESCALATE_TO_HUMAN statuses, independently triggerable stages, configurable thresholds per client, multi-client state and a central audit log.
What makes content citable by AI engines — every article is scored before it can go live.
| Criterion | Points | Check |
|---|---|---|
| Direct answer format | 15 | Clear answer to the query within the first 150 words |
| Entity density | 15 | ≥ 8 named and defined entities |
| Fact density | 15 | ≥ 5 verifiable statistics with named sources |
| FAQ section | 15 | Mandatory — ≥ 3 likely follow-up questions with direct answers |
| Structured answers | 10 | Q&A blocks, numbered lists, definition blocks |
| Schema readiness | 10 | Article/BlogPosting + FAQPage + Author JSON-LD |
| Unique perspective | 10 | An angle competitors are not using |
| Word count | 10 | ≥ 2,000 words of expert-level content |
Three gates keep specialists in control without slowing the machine down: the content plan requires specialist approval before production starts; drafts can require client review before publishing (configurable per client); the performance report requires specialist sign-off before optimization runs. Anything the agents can't resolve raises an ESCALATE_TO_HUMAN into the escalation queue.
The client-facing dashboard connects over a small authenticated REST API (per-client x-api-key):
GET /api/plugin/published → published content metadata (URL, title, keywords, date, schema) GET /api/plugin/performance → latest rankings, per-engine citations, traffic, dropped pages POST /api/plugin/settings → client settings push (keywords, niche, competitors, brand voice) POST /api/plugin/content-decision → client approvals (approved / rejected / revision requested)
The engine, orchestration, gates, scoring and APIs are fully implemented and covered by an automated end-to-end test suite. Every external service (Google Ads, DataForSEO, Search Console, GA4, Ahrefs, WordPress REST, IndexNow, the writing LLM, and per-engine citation checks) sits behind an adapter interface with deterministic mocks, so the whole system runs — and can be demoed — without credentials. Swapping a mock for a live integration is a one-file change per service.
The interactive demo runs the real agents entirely in your browser: onboard a client, run each stage, approve plans, watch the quality retry loop, publish, see a citation drop get flagged and optimized. Nothing leaves the page.