lw-elallas-integration

Pass

Audited by Gen Agent Trust Hub on Sep 9, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill identifies integration points that process untrusted data, creating a potential surface for indirect prompt injection.
  • Ingestion points: The elallas_case_status_changed hook described in SKILL.md and reference.md accepts a $message string which originates from admin inputs or system updates.
  • Boundary markers: The provided documentation and code snippets do not implement explicit delimiters or instructions for the agent to ignore embedded commands within these status messages.
  • Capability inventory: The integration logic encourages the use of network operations (syncing to external CRMs) and dynamic content generation (PDF and email updates) using the data from these hooks.
  • Sanitization: There is no mandatory requirement or instruction for sanitizing or validating the $message content before it is processed by downstream systems or interpolated into other prompts.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 9, 2026, 07:06 PM
Security Audit — agent-trust-hub — lw-elallas-integration