§ 02 · Features

The full AEO stack,
in one product.

Mentionwell is not an AI writer. It’s a measurement-led AEO platform: 10 answer engines tracked, every citation explained, an Auto Agent that ships the articles that close the gap, and a headless delivery layer that drops into any CMS or framework.

CAPABILITIES ·AEO scanning ·Pipeline ·Auto Agent ·Headless delivery ·Workspaces ·Cost console ·Indexing ·Import / Export ·6 languages (incl. RTL) ·MCP & CLI
§ 02.01 · AEO scanning

Ten engines.
Continuously measured.

Every onboarded site gets a continuous AEO scan across up to 10 engines. We capture answers, citations, fan-out queries, share of voice, sentiment, and the AI bots actually crawling your pages. Read the full AEO platform →

01

10 engines

ChatGPT, Claude, Perplexity, Gemini, Google AI Overviews, Google AI Mode, Microsoft Copilot, Grok, Exa, DeepSeek.

02

Prompt corpus

Auto-generated buyer-intent prompts from your site, ICP, value props, and competitors. Editable in Prompt Explorer.

03

Fan-out queries

Capture every sub-query each engine searched before composing the answer. Becomes article H2s and FAQs.

04

Citations table

Every cited URL — domain, title, snippet, position, owned/competitor/third-party — searchable, exportable.

05

Page evidence

Cited pages scraped: schema, headings, byline, statistics, claim-level facts. You see exactly what won the citation.

06

Share of voice

Brand mentions trended by engine, locale, topic cluster, prompt, and competitor. Proves AEO is moving.

07

Sentiment

Per-citation sentiment so you spot when engines cite you in a critical context.

08

Bot crawls

Server-log AI-bot visits — GPTBot, ClaudeBot, PerplexityBot, Google-Extended, Bingbot — per site.

09

Recommendations

Top-5 actions per scan: prompts to defend, competitor pages to out-cite, articles to write next.

§ 02.03 · Auto Agent

The blog runs itself.
When you want it to.

Set headlines per week and the days/UTC hours you want articles to ship. The Auto Agent plans the topic queue from AEO recommendations, drafts on cadence, runs the critic, generates 4K images, and publishes automatically — holding only thin-research or low-confidence drafts for your review.

01

Headline planning

Auto-generates a topic queue from the site brand profile, AEO scan recommendations, and your headlines-per-week setting.

02

Scheduled scans

Runs daily or weekly AEO scans on the cadence your plan supports. Recommendations refresh automatically.

03

Drafting

Picks the next topic, runs the multi-stage drafting pipeline (with or without Run + AEO), generates images, runs the critic.

04

Approval gates

Stay manual at any stage — the kanban (Pipeline / Drafts / Published) lets a human approve each step, or none.

05

Publish + revalidate

Pushes the article to your CMS or fires the HMAC-signed webhook so your site revalidates instantly.

06

Quota-aware

Respects your monthly article + prompt-run budget. Stops cleanly at the cap, sends a quota-80 / quota-100 email.

§ 02.04 · Optimization layers

AEO leads.
Plus GEO, LLMO, classic SEO.

Every article ships with all four optimization layers. AEO is primary; GEO, LLMO, and classic SEO are layered underneath so you don’t lose blue-link traffic while the answer-engine surface grows.

AEO [01]

Answer Engine Optimization

Question-led headings, TL;DR block, FAQ accordion, citation list. The full AEO answer surface on every article.

GEO [02]

Generative Engine Optimization

Optional Run + AEO injects observed prompts, fan-out, citations, claims, and competitor gaps into draft, metadata, and FAQ stages.

LLMO [03]

LLM Optimization

llms.txt + llms-full.txt on the parent site; markdown-source endpoints (/post.md) for AI crawlers; explicit AI bot allowlist.

SEO [04]

Classic SEO

JSON-LD BlogPosting + BreadcrumbList, sitemap.xml, RSS / JSON Feed, canonical URLs, OG tags, image alt text, internal linking.

§ 02.05 · Headless delivery

You choose how
articles land.

Pull, push, or commit — Mentionwell ships to wherever you already publish. Three live delivery modes on every paid plan (Reader API, signed webhook, GitHub MDX), with direct per-CMS push rolling out.

01

Reader API

Pull articles from a public bearer-token-scoped endpoint via the mentionwell-reader npm package or raw REST.

GET /v1/articles
02

Webhook push

Mentionwell POSTs every published article to your endpoint with HMAC-SHA256 signing.

POST → your.site/hook
03

GitHub MDX

Mentionwell commits MDX files into your repo for static-build pipelines (Astro, Next, Hugo, Eleventy).

git commit → /content
04

Direct CMS push

Native push into WordPress, Webflow, Ghost, Shopify, and Notion via official APIs is rolling out per-CMS; today these are reached through the signed-webhook adapter path (the WordPress plugin receives it natively).

WP / Webflow / Ghost…
Direct CMS push targets · rolling out

WordPress

Push into wp-admin via the REST API + Application Passwords. No plugin, no shortcode required.

Webflow

Push into a Blog Posts CMS Collection through the Webflow Data API v2.

Ghost

Push via the Admin API with Lexical-aware HTML conversion for Ghost’s editor.

Shopify

Push into Shopify’s built-in blog via the Admin API /blogs/{id}/articles endpoint.

Notion

Push into a Notion database via the official API. Render with Super.so / Potion / Feather.
§ 02.06 · Brand profiles

A profile per site,
read on every draft.

Every site you onboard gets its own brand profile. The pipeline reads from this profile on every draft so a cottage real-estate site never gets a tech-startup hero image — and a competitor blocklist actually prevents them from being mentioned.

  • Tone, audience, pitch rules.
  • Competitor blocklist + alternative-comparison list.
  • Image style brief — claymation, editorial photo, studio tech, cinematic portrait.
  • Hand-tuned taxonomy of content clusters.
  • Per-site llms.txt + AI crawler allowlist.
§ 02.07 · Image pipeline

Visual identity,
consistent at scale.

Image generation respects a per-site imageStylePrompt. Featured + inline images at 4K, mirrored to Cloudflare R2, regenerable post-publish.

  • Per-site style brief
  • 4K resolution
  • R2 mirroring
  • Post-publish regen
  • Inline + featured
§ 02.08 · Workspaces & aggregate

One dashboard.
Hundreds of sites.

For agencies and SaaS portfolios. Run hundreds of sites from one dashboard with per-site reader keys, quotas, image styles, and a workspace switcher with role-scoped seats.

  • Run hundreds of sites from one dashboard.
  • Per-site reader keys, quotas, and image styles.
  • Aggregate dashboard with cross-site article + scan health.
  • Workspace switcher with role-scoped seats (Growth+).
  • White-label reader package for agency-deployed sites.
  • Admin console for super-admins: cost ledger, queues, balances, plan grants.
§ 02.09 · Cost console

Every token.
Ledgered.

Most AI blog tools won’t show you their token spend. We do — per article, per stage, per model, per site. Exportable, auditable, ungated.

01

Per-article cost — input + output tokens, model, total USD.

02

Per-stage cost — research, draft, critic, metadata, embedding, images.

03

Per-model cost — every OpenRouter / FAL call ledgered.

04

Per-site rollup — what each domain in your portfolio is costing.

05

Daily, 7d, 30d, all-time totals — exportable to CSV.

06

Admin console: cross-tenant cost ledger for super-admins.

§ 02.10 · Indexing & analytics

Indexed in seconds.
Measured forever.

Publish triggers IndexNow, Bing Webmaster, and the Google Indexing API where eligible. Search Console + GA4 OAuth-connect so per-article performance flows back into the planner.

01

Google Search Console

OAuth-connect GSC. Pull queries, clicks, impressions, average position per page. The dashboard surfaces winners + losers per article.

02

Google Analytics 4

OAuth-connect GA4. Per-article sessions, engaged sessions, bounce rate. Quality signal back into the planner.

03

IndexNow

Automatic IndexNow ping on publish — Bing, Yandex, Naver pick up the article in seconds.

04

Bing Webmaster Tools

Bing index API + crawl-stats integration.

05

Google Indexing API

Job-posting + livestream URLs auto-submitted; non-eligible URLs fall back to sitemap ping.

06

PageSpeed Insights

Per-article Core Web Vitals + LCP/CLS/INP report. Image regen targets the worst LCP.

§ 02.11 · Import / Export

Bring an archive.
Leave anytime.

Import from any major platform — slugs preserved, rankings intact. Then bulk-optimize for AEO. And every article is exportable as Markdown, WXR, or a zip with images. You own your content.

Import sources

  • WordPress (XML or REST)
  • Webflow
  • Ghost
  • Substack
  • RSS feed
  • Sitemap.xml
  • Manual paste

Export formats

  • Markdown (per article)
  • WXR (WordPress XML)
  • Zip bundle (Markdown + images)
  • Image bundle (R2-backed)
  • JSON snapshot
§ 02.12 · Developer surface

Headless API.
Plus MCP, CLI, SDK.

The full programmatic surface: read articles, push articles, fire webhooks, and let your AI coding agent talk to Mentionwell directly via MCP.

01

Reader API

Public, bearer-token-scoped /v1/articles + /v1/articles/{slug}. JSON-clean output ready for any framework.
02

Reader SDK

mentionwell-reader npm package — typed loaders for Next, Astro, Remix, Nuxt, SvelteKit, React, plain HTML.
03

Webhooks

HMAC-SHA256-signed POST on every publish event with payload schema and replay-protection headers.
04

Personal access tokens

mw_pat_… tokens scope per workspace. Power the account MCP and any other headless surface.
05

MCP servers

mentionwell-mcp-docs (no-auth, drop into Claude/Cursor/ChatGPT for accurate docs answers) + mentionwell-mcp-account (PAT-authed, list sites, run Connect, verify webhooks, import posts).
06

CLI

npx mentionwell-cli init — autodetect framework, write the revalidation route, sync env vars, verify the webhook end-to-end.
§ 02.13 · Connect flow

Onboard a site
in under five minutes.

The Connect flow is agentic: probe → confirm → migrate → execute → verify. Branded preview pages confirm each handshake. By the time you blink, a test article has round-tripped to your CMS.

01

Probe

Crawl the destination site, detect framework, CMS, existing content schema, sitemaps, RSS.

02

Confirm

Surface a one-page summary the operator confirms — domain, brand profile, taxonomy, image style.

03

Migrate

Optionally bulk-import existing posts from the detected source so AEO measurements run on a real corpus.

04

Execute

Wire delivery — webhook secret, reader key, GitHub PAT, or per-CMS credentials. Branded preview pages confirm each handshake.

05

Verify

Round-trip a test article. Webhook fires, revalidation triggers, the article is live. Then we hand off to autopilot.

§ 02.14 · Languages & locales

Six languages.
Native, not translated.

The pipeline runs natively in six tier-1 locales — English (US/UK), German, French, Spanish, Hebrew with full RTL. Per-locale guardrails on every prompt: spelling, idiom, pronouns, currency, decimals, dates, quotation marks. Hebrew gets the deepest spec — no niqqud, gershayim, Latin-script brand names. Read the full language spec →

  • 🇺🇸 en-US English (US)
  • 🇬🇧 en-GB English (UK)
  • 🇩🇪 de-DE Deutsch
  • 🇫🇷 fr-FR Français
  • 🇪🇸 es-ES Español
  • 🇮🇱 he-IL עברית · RTL
§ 02.15 · Email lifecycle

You’re notified when it matters.

Thirteen transactional templates, sent only when something useful happened. Every email respects per-user preferences and a one-click unsubscribe.

Welcome
First-run summary
Headlines ready
Article drafted
Article published
Draft failed
Weekly digest
Quota at 80%
Quota at 100%
Trial day 7
Trial day 12
Trial expired
Payment failed
§ 02.16 · See also
AEO platform Pricing Integrations Languages For agencies Docs Changelog
JOIN THE BETA

Be in the first 200 sites we onboard.

Mentionwell launches June 1, 2026. We're letting sites in week by week — drop your email and we'll reach out when your spot is ready.

// 66 sites beta testing now · early access invites going out weekly
Live ↗toloseadog.comsethdive.comitsbaba.comdrabinskyrealty.com

Prefer a walkthrough first?

66 beta testing now