Never miss what matters.

Watchflare is a Monitoring Intelligence Platform. You define a monitoring job — a topic, a set of URLs, and an intelligence context — and we automatically scrape, score, and summarize what changes. Think of it as a radar for any topic on the web.

Any publicly accessible URL: news sites, competitor pages, government publications, regulatory bodies, research papers, community forums, and more. If it's on the web and accessible, we can track it.

You provide a custom intelligence context when creating a job (e.g. 'Monitor AI startup funding rounds under $10M'). Our AI analyzes every new piece of content against that context and assigns a score from 0 to 100. Only high-scoring findings reach you.

Credits are our atomic billing unit. One scrape + one AI analysis = approximately 1 credit. If a page hasn't changed since the last scrape, we skip the analysis and charge 0 credits. Cached pages from other users are also free.

Every webhook payload is signed using HMAC-SHA256 with your account secret. The signature is sent in the X-Watchflare-Signature header. Verify it on your end before processing any event.

Watchflare ships as a Model Context Protocol (MCP) server. You can plug it into Claude Desktop, Cursor, Windsurf, or any MCP-compatible AI client. Your agent gets 10 tools to interact with your monitoring jobs, findings, and briefings in real time.

Yes. You get 100 credits on signup — no credit card required. That's enough to set up 3 monitoring jobs and run your first cycles. Upgrade when you need more volume.

Google Alerts only emails you when Google indexes a mention — which can take hours or days. Watchflare scrapes your exact URLs on a schedule you define, scores content with AI relevance analysis, and delivers structured briefings. It's built for professionals who need speed, precision, and context — not just keyword matches.

Yes. Watchflare includes a native MCP server that plugs directly into Claude Desktop, ChatGPT (via plugins), Cursor, and any MCP-compatible AI client. Your AI agent can query your live monitors, fetch findings, and compile briefings in real time — no coding required.

We use SHA-256 content hashing to detect changes. After each scrape, we hash the extracted content and compare it with the previous version. If the hash matches, we skip the AI analysis entirely — saving credits and avoiding duplicate findings.

Yes. Your monitoring jobs, intelligence contexts, and findings are fully private and isolated per account. We use Row-Level Security (RLS) on every database table. The only shared component is the content cache — cached page content (not your analysis) is shared across tenants to reduce scraping costs.

MCP (Model Context Protocol) is a standard that lets AI agents connect to external tools and data sources. Watchflare's MCP server gives your AI agent 10 tools to create jobs, fetch findings, compile briefings, and more. It means your AI assistant can actively monitor the web for you — not just answer questions about what it already knows.