Octolens已验证已精选

What it does

Octolens is one API for every mention across the internet, and this MCP server puts it directly into your AI agent's hands. Ask Claude what people said about your brand on Hacker News this week, have an agent triage feature requests from Reddit, or run SQL over your entire mention history — no dashboard, no manual export, no juggling a dozen platform APIs.

Coverage spans 15+ platforms: Reddit, X (Twitter), LinkedIn, Hacker News, YouTube, Bluesky, GitHub, Stack Overflow, dev communities, podcasts, newsletters, news, Product Hunt, Medium, and TikTok. Every mention is AI-filtered for relevance and scored for sentiment before it reaches your agent — which matters most for brands whose name doubles as a common word, where raw keyword feeds are unusable noise.

Tools

Tool	What it does
list_mentions_context	Returns filter syntax, your tracked keywords, and saved views — call first so the agent knows how to query
list_mentions	Fetch mentions with filters and pagination: by keyword, platform, sentiment, relevance, date, tags
analytics_context	Returns the analytics schema and SQL tips for the mentions database
analytics	Run ClickHouse-compatible SQL over your full mention history — trends, share of voice, sentiment breakdowns

The context tools mean agents self-orient: no prompt engineering needed to teach them your filter syntax or schema.

Example prompts

• "What are people saying about us on Reddit and Hacker News this week? Summarize sentiment."
• "Find every mention tagged as a feature request this month and group them by theme."
• "Compare our mention volume against [competitor] over the last quarter, by platform."
• "Which negative mentions from the past 48 hours should I respond to first?"
• "Pull five positive quotes about our API for a landing page."

Who it's for

Brands people talk about online — developer tools, product-led SaaS, and companies with active communities. Teams like Vercel, PostHog, and Prisma use Octolens to catch conversations across the web; the MCP server is how that data reaches agent workflows: support agents that know what users complain about, research agents with real voice-of-customer data, and marketing agents that pull social proof on demand.