MCP servers for marketing automation, SEO, email campaigns and social-media management. Each entry is scanned with the deterministic Capability-Flow Trust Model — grades are computed, never self-reported.
MCP Server for developers building HubSpot Apps
MCP (Model Context Protocol) server that helps AI agents use Primer Brand (@primer/react-brand) correctly when building GitHub marketing and landing pages.
Operational empathy for technical founders + PE/VC operators navigating high-stakes B2B interactions. 28 MCP tools + CLI: ICP scoring, persona simulation, battlecards, deal classification, prospect discovery, investor matching, founder wellness, plus mark
Model Context Protocol (MCP) server that exposes 100% of eBay's Sell APIs (322 tools across 270 endpoints) to AI assistants like Claude, Cursor, and Cline, covering inventory, order fulfillment, marketing, analytics, and developer tools with OAuth authent
Official Model Context Protocol (MCP) server for the Kaminari Ad ad-verification platform.
LinkedIn MCP server — structured LinkedIn data via a real stealth-browser session. Profiles, search, jobs, companies, messaging, network. Works with Claude, Cursor, and any MCP client.
Read-only MCP server for VR.org. Gives AI agents one-call access to live VR, AR, and XR news, original editorial with full article text, the events calendar, headset deals, buyer guides, and top-game and top-app lists.
MCP server for Kadam ad network — manage campaigns, creatives, audiences, sites, and analytics via AI agents
SFMC AMPscript/SSJS/GTL tools + searchable MCE and MCN developer docs, with full MCN platform support.
LinkedIn MCP server — publish posts, articles, and media via the LinkedIn API
Lemmy federated link-aggregator public reads (any instance)
User-token (xoxp) Slack MCP server over stdio
MCP server for Bing Webmaster Tools — SEO site management, URL submission, sitemaps, stats, keywords
Typed MCP server for OpenAI Ads and ChatGPT Ads via the Advertiser API
Model Context Protocol server for svelte-vitals — run SEO analysis inside an agent's tool loop.
Community MCP server for the Peec AI API - AI Search Analytics
MCP server for AI-powered brand identity generation — color palettes, typography systems, design tokens, and brand guidelines with multi-format export
MCP server for Nestr - the platform for self-organizing teams practicing Holacracy, Sociocracy, and role-based work. Manage circles, roles, tactical meetings, governance meetings, tensions, and more.
Fuul MCP server for program management via MCP-compatible clients
MCP server for Plausible Analytics — query your website stats from AI assistants like Claude
Mailchimp MCP Pack — manage audiences, campaigns, and members via Mailchimp Marketing API.
Mastodon MCP — public Mastodon data via mastodon.social (no auth required)
Reddit MCP — public Reddit data via JSON endpoints (no auth required)
Slack MCP Pack
Growth work already runs on APIs: social platforms (X, LinkedIn, Instagram, Facebook, Reddit), email services like Mailchimp, SendGrid and HubSpot, SEO data from Google Search Console, Ahrefs or Semrush, and ad accounts on Google Ads and Meta. A marketing and social MCP server wraps those APIs as structured tools any Model Context Protocol client — Claude, Cursor, other LLM assistants — can call directly. Instead of summarizing dashboard screenshots, the assistant reads and writes campaigns, audiences and analytics itself.
Few categories combine a public voice with private customer data like this one. An over-permissioned or compromised server could publish under your brand handles, mail your whole subscriber base — phishing from a verified sending domain — or exfiltrate contact lists full of personal data. Prompt injection is a live path too: comments, mentions, DMs and inbound email are attacker-controlled text that lands in the assistant's context before any write action fires. That exposure is why every listing in the registry carries an A–F Trust Score from a deterministic open-source engine: it parses the published package, maps declared capabilities into a blast-radius profile, and records each security finding with evidence — identical input always grades identically, and placement cannot be bought. The scoring model is documented under how the algorithm works. Two neighbouring categories overlap here: creator and streaming promotion borders Gaming & Entertainment, while attribution and audience modeling leans on Data Science & ML.
They can be, if treated like any third-party integration with posting rights. Check the Trust Score and read the individual findings, confirm the blast radius stays inside social and email APIs — a posting tool has no business requesting shell or filesystem access — and scope tokens to only the accounts and permissions a workflow needs. Keep inbound social content classified as untrusted input before it triggers any write to a brand channel.