skills/mukul975/anthropic-cybersecurity-skills/implementing-zero-trust-dns-with-nextdns/Gen Agent Trust Hub
implementing-zero-trust-dns-with-nextdns
Fail
Audited by Gen Agent Trust Hub on Apr 10, 2026
Risk Level: CRITICALEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill fetches the official NextDNS installer script from
https://nextdns.io/install. This is the standard deployment method provided by the service for its CLI agent. - [COMMAND_EXECUTION]: Utilizes
sudofor administrative tasks such as modifying/etc/systemd/resolved.confand installing the NextDNS service agent, which is necessary for changing system-level network configurations. - [DATA_EXFILTRATION]: Communicates with the official NextDNS REST API (
api.nextdns.io) to retrieve account profile settings and DNS query logs for auditing purposes. - [PROMPT_INJECTION]: The skill processes untrusted DNS log data, creating an indirect prompt injection surface.\n
- Ingestion points: DNS query logs retrieved from the API in
agent.pyand sample logs inprocess.py.\n - Boundary markers: Absent from log processing logic.\n
- Capability inventory: Includes system-level configuration commands and script execution capabilities.\n
- Sanitization: No validation or escaping is performed on external domain names from the logs before they are processed or displayed.
Recommendations
- Contains 1 malicious URL(s) - DO NOT USE
Audit Metadata