Sends payment links on behalf of users — but never a phishing URL
WHAT BEOMZ PROTECTS
Every URL extracted from drafts or pasted by the user is verified against phishing databases and domain reputation signals before the email is dispatched.
DEFAULT INPUT
Phishing payment URL
A lookalike domain targeting a SaaS billing page — classic credential-harvest phishing.
WALLET MODE
SUBJECT TO VERIFY
LIVE VERDICT
No verdict yet
Run the demo above to see a live Beomz verdict.
FRAMEWORK-IDIOMATIC INTEGRATION
CrewAI# CrewAI agent with Beomz URL verification
from beomz_crewai import BeomzVerifyUrlTool
from crewai import Agent, Task, Crew
verify_url_tool = BeomzVerifyUrlTool(
network="testnet",
account_id=os.environ["BEOMZ_DEMO_ACCOUNT_ID"],
private_key=os.environ["BEOMZ_DEMO_PRIVATE_KEY"],
)
security_agent = Agent(
role="Email Security Guard",
goal="Block phishing links before emails are sent",
tools=[verify_url_tool],
backstory="I scrutinize every URL in outgoing emails.",
verbose=True,
)
verify_task = Task(
description=(
"Verify this URL before sending: "
"https://billing.somedomain.tld/pay/123. "
"If Scam or Risky, halt and alert the user."
),
agent=security_agent,
expected_output="Verdict with block/allow decision",
)
crew = Crew(agents=[security_agent], tasks=[verify_task])
result = crew.kickoff()