sentry

Warn

Audited by Snyk on May 4, 2026

Risk Level: MEDIUM
Full Analysis

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

  • Third-party content exposure detected (high risk: 0.70). The SKILL.md explicitly instructs using the Membrane CLI to connect to Sentry and run actions like "list-issues" and "get-issue", which fetch user-generated Sentry data (issue titles/bodies, comments, etc.) that the agent is expected to read and could materially influence follow-up actions (e.g., update/delete), exposing it to untrusted third-party content.

Issues (1)

W011
MEDIUM

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

Audit Metadata
Risk Level
MEDIUM
Analyzed
May 4, 2026, 03:48 PM
Issues
1
Security Audit — snyk — sentry