url-get-gti-report

name	url-get-gti-report
description	Use when retrieving GTI threat classification and URL category intelligence.
category	atomic
version	1.0.0
type	Skill
title	Skill: Atomic Runbook: Get URL Reputation from GTI
generated	{"by":"process:google-labs-jules","at":"2026-08-20T02:00:00Z"}

Atomic Runbook: Get URL Reputation from GTI

ID: RB-ATOM-URL-001 Version: 1.0 Last_Updated: 2025-05-30 Purpose: To retrieve a comprehensive URL analysis report from Google Threat Intelligence (GTI) to assess its reputation, categorization, and any associated threats. Parent_Runbook(s)/Protocol(s): rules-bank/indicator_handling_protocols.md#url-indicator (Assuming a future section for URLs) Trigger: When a URL requires an external reputation check as part of an investigation (e.g., from a phishing email, web proxy logs, or EDR alert).

---

Inputs Required

• url: string - The full URL to query (e.g., http://example.com/path/to/file.html).
  • Source Example: Alert field target.url, email body, web proxy logs.

---

Execution Steps

1. Tool Selection:
   • Primary_Tool_MCP_Server: Google Threat Intelligence MCP
   • Primary_Tool_Name: get_url_report
2. Parameter Mapping:
   • Map url (Input) to MCP Tool parameter url.
3. Execute Tool: Call the get_url_report tool with the mapped url.
   • AI Agent Note: Refer to rules-bank/mcp_tool_best_practices.md for specific guidance on GTI tools. Ensure URL is properly encoded if necessary, though the tool should handle standard cases.
4. Data Transformation/Extraction:
   • The primary output is a JSON object. Key fields for assessment include data.attributes.last_analysis_stats (especially malicious), data.attributes.categories, data.attributes.final_url (after redirections), and data.attributes.redirection_chain.

---

Outputs Expected

• gti_url_report: JSON - The full JSON report from the GTI get_url_report tool.
• malicious_score: integer - Engines reporting the URL as malicious (from gti_url_report.data.attributes.last_analysis_stats.malicious).
• suspicious_score: integer - Engines reporting as suspicious.
• harmless_score: integer - Engines reporting as harmless.
• categories: list - List of categories assigned to the URL by GTI (e.g., "phishing", "malware", "benign").
• final_url: string (optional) - The final URL after any redirections.
• redirection_chain: list (optional) - List of URLs in a redirection chain.
• output_status: string - ["Success", "Failure", "NotFound"]
  • "NotFound" if GTI has no information on the URL.
• output_message: string (if Failure) - Details of the issue.

---

Decision Logic / Next Steps (If Applicable)

• IF output_status is "Success":
  • IF malicious_score > 5 OR "phishing" IN categories OR "malware" IN categories THEN
    • Flag URL as "High_Risk_GTI".
    • Extract domain from url (or final_url) and consider initiating RB-ATOM-DOMAIN-001 (Get Domain Reputation From GTI).
    • Consider proceeding to containment (e.g., block URL/domain) or escalate immediately.
  • ELSE IF malicious_score > 0 OR suspicious_score > 0 THEN
    • Flag URL as "Medium_Risk_GTI_Needs_Correlation".
    • Extract domain and proceed to rb_url_search_chronicle.md and domain-specific runbooks.
  • ELSE (low malicious/suspicious score)
    • Flag URL as "Low_Risk_GTI".
    • Proceed to rb_url_search_chronicle.md for internal context.
• IF output_status is "NotFound":
  • Log "URL {url} not found in GTI."
  • Flag URL as "Unknown_Reputation_GTI".
  • Extract domain and proceed with domain reputation checks (RB-ATOM-DOMAIN-001) and internal searches (rb_url_search_chronicle.md).
• ELSE (output_status is "Failure"):
  • Log error: output_message.
  • Consider executing rb_url_get_secops_threat_intel.md as an alternative.
  • IF alternative also fails, escalate: "Failed to retrieve external reputation for URL {url}."

---

AI Agent Execution Notes

• The AI should parse the gti_url_report to extract specific output parameters.
• If a final_url is present and different from the input url, the AI should note this and potentially perform checks on the final_url as well.
• The domain of the URL should be extracted for separate domain reputation checks.

---

Metrics Collection Points

• Log execution time.
• Log output_status, malicious_score.
• (Reference rules-bank/ai_performance_logging_requirements.md)

---

References

• rules-bank/mcp_tool_best_practices.md
• rules-bank/indicator_handling_protocols.md (future URL section)
• rules-bank/ai_performance_logging_requirements.md