halo-change-request

Warn

Audited by Snyk on Jul 30, 2026

Risk Level: MEDIUM
Full Analysis

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

  • Third-party content exposure detected (medium risk: 0.30). The workflow ingests outsider-authored free text when it reads existing Halo tickets as a “real example” via halo_list_tickets/halo_get_ticket, and those ticket contents can include user-generated text that the LLM will process during field learning.

Issues (1)

W011
MEDIUM

Third-party content exposure detected (indirect prompt injection risk).

Audit Metadata
Risk Level
MEDIUM
Analyzed
Jul 30, 2026, 04:29 PM
Issues
1
Security Audit — snyk — halo-change-request