The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes various system-level profiling and management commands including network port scanning (ss, lsof), firewall status checks (ufw, firewall-cmd, nft, socketfilterfw), and system version checks.
[COMMAND_EXECUTION]: The skill is designed to modify host security configurations, such as firewall rules, SSH/RDP settings, and system service policies upon user approval.
[COMMAND_EXECUTION]: The skill uses the openclaw cron utility to establish persistent scheduled tasks on the host for periodic security audits and update checks.
[DATA_EXFILTRATION]: The skill performs extensive reconnaissance on the host environment, collecting data on OS versions, privilege levels, network exposure, disk encryption status, and backup configurations.
[DATA_EXFILTRATION]: The skill writes detailed summaries of actions and system state to local memory files (memory/YYYY-MM-DD.md), which may include sensitive host metadata.
[EXTERNAL_DOWNLOADS]: The skill queries the NPM registry to check for software versions using the command npm view openclaw version.
[PROMPT_INJECTION]: The skill contains an indirect prompt injection surface by ingesting and processing untrusted data from system tool outputs and external security audits without explicit boundary markers or sanitization logic.
Ingestion points: SKILL.md (Workflow Step 1 & 2) ingests data from system commands (uname, ss, ufw) and openclaw security audit.
Boundary markers: Absent; there are no instructions to the agent to ignore or delimit embedded instructions within the processed tool outputs.
Capability inventory: The skill has high-privilege capabilities including system configuration modification and persistent task scheduling via openclaw cron.
Sanitization: Absent; the instructions do not specify validation or filtering of the ingested content before processing.

healthcheck