Analyze SMS phishing with the AI action

Danielle Swanser

Danielle Swanser

Product Engagement Manager

Security

Employees report suspicious text messages via a Tines Page, and the AI action analyzes the screenshot to extract indicators and deliver a verdict — while automatically creating a case, enriching IOCs, and escalating if multiple employees report the same threat.


What you'll see:

  • Employees submit a screenshot of a suspicious text message through a branded Tines Page
  • An LLM action powered by Claude Sonnet analyzes the image and returns a verdict (malicious, spam, or undetermined), confidence score, extracted URLs, and sender phone number
  • Condition actions validate the AI output format and data types, looping back into a corrective LLM action if the response doesn't meet expectations
  • A Jira issue is created for the SOC, with the screenshot attached
  • Extracted URLs are scanned in parallel using URLScan and VirusTotal via Send to Story actions
  • The extracted phone number is validated, queried in C99, and added to MISP as a threat indicator
  • Enrichment results are posted back as comments on the Jira issue
  • Historical reports stored in a Tines Resource are checked for matching URLs or phone numbers across previous submissions
  • If matches are found, a Slack alert is sent to the SOC channel and a PagerDuty incident is created
  • A personalised response is built and displayed to the employee via a Thank You page

Perfect for:

  • SOC teams: Triaging employee-reported SMS phishing at scale without manual intake or analyst involvement
  • Security awareness programs: Giving employees a fast, visible feedback loop when they report suspicious messages
  • Threat intelligence teams: Automatically surfacing and correlating IOCs (phone numbers, URLs) across multiple employee reports
  • Incident response teams: Identifying coordinated phishing campaigns early when multiple employees receive the same malicious message

Requirements:

  • Tines AI enabled on your tenant
  • Anthropic Claude Sonnet (via Tines AI — model: (claude-3-sonnet-20240229-v1:0)
  • Slack — Bot Token with chat:write scope
  • PagerDuty — API key with permissions to create incidents
  • Jira — API key and a service account with permission to create issues and add attachments
  • C99 — API key for phone number lookup (api.c99.nl)
  • MISP — API key for adding threat indicators
  • Tines API key — for reading and writing to a Tines Resource used to store historical reports
  • A URLScan Send to Story and a VirusTotal Send to Story configured and available in your tenant
  • Company name, domain URL, and relevant keywords (e.g. company name, CEO name) added to the Build Company Details and Build Keywords actions before use

Get the featured workflow

Import this workflow to your tenant, from where you can adapt it to meet your unique needs.

View workflow
🔎✉️✨Analyze SMS phishing with the AI actionTools: MISP, Slack, URLScan.io, VirusTotal

Receive Five-minute flows
directly in your inbox

Built by you,
powered by Tines

Already have an account? Log in.

Analyze SMS phishing with the AI action | Tines