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
MEDIUMThird-party content exposure detected (indirect prompt injection risk).
Audit Metadata