openclaw-ops-expert

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

• Third-party content exposure detected (high risk: 1.00). The skill explicitly describes integrating with public messaging platforms (e.g., Telegram/Slack/Discord/WhatsApp) and shows message flow/gateway ingestion in SKILL.md and references (see "Channel Integration" in references/overview.md, the messages flow diagram, and the Telegram Channel Integration workflow in references/scenarios.md §9.2), meaning the agent will read untrusted, user-generated third‑party messages that can influence its actions.

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

• Attempt to modify system services in skill instructions detected (high risk: 0.80). This skill explicitly directs creating and managing system daemons and system-level service files (systemd unit files, launchd plists, "openclaw onboard --install-daemon", npm -g installs and daemon onboarding) and promises copy-paste commands, which involve modifying system files and likely require root/sudo access, so it poses a significant risk of changing machine state.