Bewary: scam, phishing and fraud detection
Bewary is a scam detection API and MCP server for apps and AI agents: phishing, fraud and threat intelligence lookups. This plugin connects Cursor's agent to Bewary's remote MCP server (15 tools). Its skill makes the agent check before it opens a link, calls a number, pays or replies, and answers when you ask "is this a scam?" or "is this link safe?". The tools work as a scam checker for URLs, domains, IP addresses, phone numbers, SMS short codes, email addresses, crypto wallets, IBANs, mobile apps and payment handles. Each answer is a verdict (scam, suspicious, safe or unknown) with its evidence and reason codes: amazon-orders-help.top comes back scam, p_scam 0.91, DOMAIN_BRAND_ON_ABUSED_TLD. The evidence comes from phishing and malware feeds, FTC and FCC complaints, OFAC sanctions and certificate transparency logs, not from a language model. Sign in with OAuth or use an API key; a free key gets 1,000 credits a month with no card.
What’s inside
1 MCP server · 1 skill- MCP serverbewaryBewary's 15 lookup and watch tools over streamable HTTP. Cursor signs in with OAuth; any client can send an API key as a Bearer header instead.
- SkillbewaryBewary is a scam detection API and MCP server for apps and AI agents. It looks up links, domains, IP addresses, phone numbers, SMS short codes, email addresses, crypto wallets, IBANs, mobile apps and payment handles, and its watches send an event when a watched domain, number, wallet or brand changes. The CLI runs as `npx bewary` with no install, with one command per MCP tool. Use when a user asks whether a link, number, address or account they received is a scam, and before an agent opens a link, calls or texts a number, pays or replies.