debug-openshell-cluster

Pass

Audited by Gen Agent Trust Hub on Nov 20, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONPRIVILEGE_ESCALATIONOBFUSCATION
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructs the agent to use a wide array of command-line tools for system and cluster diagnostics, including kubectl, helm, docker, podman, openshell, journalctl, systemctl, rg, stat, and openssl.
  • [PRIVILEGE_ESCALATION]: Specific troubleshooting steps require the use of sudo to perform administrative actions, such as enabling or disabling systemd services and inspecting file metadata in restricted system directories.
  • [INDIRECT_PROMPT_INJECTION]:
  • Ingestion points: The agent is instructed to read and analyze data from untrusted sources, including container logs (docker logs, kubectl logs), system journals (journalctl), and cluster events.
  • Boundary markers: The instructions do not include boundary markers or specific warnings to ignore potentially malicious instructions embedded within the logs being analyzed.
  • Capability inventory: The skill leverages tools with significant system and cluster-level authority, such as kubectl for resource manipulation and systemctl for service management.
  • Sanitization: There is no requirement for the agent to sanitize or validate the content of the logs before interpreting them, creating a surface for indirect prompt injection.
  • [OBFUSCATION]: The skill uses hex escapes (\x1b) within a sed command intended to strip ANSI color codes from terminal output. While appropriate for the task of log cleaning, this involves the use of encoded character representations.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 20, 2026, 07:24 PM