networking
Pass
Audited by Gen Agent Trust Hub on Apr 7, 2026
Risk Level: SAFEREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [REMOTE_CODE_EXECUTION]: Fetches and executes the acme.sh installation script from its official domain (https://get.acme.sh). This installation pattern is a standard method for the tool but bypasses traditional package verification.
- [COMMAND_EXECUTION]: Provides instructions for managing network infrastructure using standard system tools like nft, systemctl, wg, tailscale, and step. These tools are used within their intended administrative scope for network configuration and hardening.
- [PROMPT_INJECTION]: The skill exhibits a surface for indirect prompt injection because it involves the ingestion and analysis of external data sources such as network logs.
- Ingestion points: The agent is instructed to read and process network logs (e.g., /var/log/suricata/fast.log) and authentication logs to identify security threats as described in references/security-hardening.md.
- Boundary markers: There are no specified delimiters or instructions to ignore potential commands or instructions embedded within the ingested log entries.
- Capability inventory: The skill utilizes powerful system tools across multiple scripts, including firewall configuration (nft), service management (systemctl), and VPN administration (wg).
- Sanitization: No sanitization, escaping, or validation routines for the ingested log data are specified before the agent processes the content.
Audit Metadata