The Agent Skills Directory

[PROMPT_INJECTION]: The skill processes untrusted external data in the form of DNS records (TXT, MX, etc.). While this presents a surface for indirect prompt injection, it is inherent to the skill's purpose of DNS debugging.
Ingestion points: Output from DNS query tools like dog, dig, and nslookup (SKILL.md).
Boundary markers: Absent; there are no explicit instructions for the agent to treat record content as untrusted data.
Capability inventory: The skill has access to shell command execution (Bash) and file operations (Read, TodoWrite).
Sanitization: Absent; no filtering or validation of the DNS record content is performed before processing.
[EXTERNAL_DOWNLOADS]: The skill suggests the installation of the dog DNS utility, which is a well-known, open-source Rust-based tool. Installation is recommended through official and standard package managers such as cargo, brew, and pacman.
[SAFE]: The skill utilizes well-known and trusted DNS resolvers provided by Google (dns.google), Cloudflare (cloudflare-dns.com), and Quad9 (dns.quad9.net) for its encrypted DNS query examples.

dns-tools