cloudflare-dns
Fail
Audited by Gen Agent Trust Hub on Nov 13, 2026
Risk Level: HIGHEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill fetches a binary executable from a non-official GitHub repository (
wsvn53/flarectl) instead of a verified or official vendor source. - [REMOTE_CODE_EXECUTION]: The skill downloads an external binary file directly into a system path (
/usr/local/bin/flarectl), modifies its permissions to be executable (chmod +x), and subsequently runs it within the environment. - [COMMAND_EXECUTION]: Utilizes shell commands and Python scripts that invoke
subprocess.runto manage DNS records and execute system-level logic. - [INDIRECT_PROMPT_INJECTION]:
- Ingestion points: The skill reads and parses data from Cloudflare DNS records via the
flarectl dns listcommand. - Boundary markers: None present; external DNS record content is processed without delimiters or safety warnings.
- Capability inventory: The skill includes logic to perform batch deletions using
subprocess.runbased on the content of the ingested DNS records. - Sanitization: No sanitization or validation is performed on the record IDs or contents before they are passed into subsequent shell commands.
Recommendations
- AI detected serious security threats
Audit Metadata