caspio

Warn

Audited by Snyk on Apr 22, 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's SKILL.md instructs the agent to run actions and to proxy arbitrary Caspio API endpoints (e.g., "membrane request CONNECTION_ID /path/to/endpoint"), which causes the agent to fetch and interpret external Caspio data (potentially user-generated/untrusted) as part of its workflow.

Issues (1)

W011
MEDIUM

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

Audit Metadata
Risk Level
MEDIUM
Analyzed
Apr 22, 2026, 01:13 PM
Issues
1