Scam Detector
Screens messages for scams and returns an evidence-based verdict — for any AI agent that reads SKILL.md.
Free
Works with the AI Agents you already use
Scam Detector
Example session with this skill installed
Check this email I just got. It says it is from Netflix and says my payment failed, but the sender address is netflix-support@gmail.com and the link goes to bit.ly/fix-my-billing.
- Read your context and instructions
- Compiled the scam detector
I have analyzed the provided message content and link metadata.
Verdict: CONFIRMED SCAM
Confidence: 98%
Category: Credential Harvesting / Billing Phishing
Evidence Audit
| Check | Result | Details |
|---|---|---|
| Sender Identity | FAILED | Netflix uses @netflix.com or @mailer.netflix.com. The use of a free @gmail.com address for official billing is a 100% mismatch. |
| Link Destination | MALICIOUS | The destination bit.ly/fix-my-billing is a URL shortener used to hide a non-Netflix domain. Official billing updates never use shorteners. |
| Social Layer | SUSPICIOUS | Uses a "Payment Failed" lure to create artificial urgency and fear of service loss. |
| Authentication | UNVERIFIED | Pasted content; SPF/DKIM/DMARC headers were not available for alignment testing. |
Detailed Findings
The message follows a standard template for credential theft. By masking the final destination behind a Bitly link and using a generic Gmail account, the attacker bypasses simple keyword filters. The request for billing "fixing" is designed to capture your credit card details and Netflix login credentials.
Next steps
- Do not click the link or reply to the email.
- Mark as Phishing in your email client to train your local filters.
- Verify via official app: Log in directly to Netflix.com or use the official app to check your account status.
Connects securely to your tools. The creator never sees your data.
About this skill
Free, open-source agent skill (SKILL.md) that screens messages — email, WhatsApp, Messenger, Instagram, SMS, pasted text — for scams. Checks sender authentication (SPF/DKIM/DMARC and alignment), lookalike domains, links and QR codes analysed as text (never visited), social-engineering patterns, a social-scam layer (romance, pig-butchering, impersonation, job scams), and attachment metadata. Returns a graded verdict: CONFIRMED SCAM, LIKELY SCAM, SUSPICIOUS — verify independently, LIKELY GENUINE, or UNVERIFIED, with the evidence shown and any unchecked items listed. Strictly read-only: never clicks, opens, or replies. Works with Claude, ChatGPT, Grok, Muse and other SKILL.md hosts. Singapore-seeded reference data; the procedure works anywhere.
Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for Security.
Read-only static analysis. Cannot bypass encrypted traffic or private chats without user-provided exports or host-level access. Reputation checks depend on available third-party API keys.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before buying a skill.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Free to use
No account required to browse
Trust & safety
Security scan
Security scan passed 8 days ago
- Free with an Agensi account