debug-openshell-cluster

Warn

Audited by Snyk on Nov 20, 2026

Risk Level: MEDIUM
Full Analysis

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

  • Third-party content exposure detected (low risk: 0.10). The workflow reads sandbox logs, tool-server status (last_result, HttpResponseReceived text), and workload output via openshell logs, which can carry text influenced by external tool servers or workload traffic. The agent selects a specific sandbox by name and the outsider cannot push content into a feed the workflow monitors, so this is on-demand retrieval rather than forced delivery.

MEDIUM W013: Attempt to modify system services in skill instructions.

  • Attempt to modify system services in skill instructions detected (low risk: 0.10). The skill contains state-changing operational branches: enabling and restarting systemd units (lines 335-337), sudo systemctl disable --global for the snap gateway (line 76), kubectl delete statefulset --cascade=orphan followed by helm upgrade (line 1669), pg_terminate_backend on gateway lock sessions, and deletion of managed ConfigMaps. None of these is tied to an explicit, unrecoverable loss of operational state. The orphaned PVC is retained, the Podman cache is rebuilt, and the sandbox-deletion workspace warning concerns user workspace content rather than operational capability, so the tier is low.

Issues (2)

W011
MEDIUM

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

W013
MEDIUM

Attempt to modify system services in skill instructions.

Audit Metadata
Risk Level
MEDIUM
Analyzed
Nov 20, 2026, 07:24 PM
Issues
2