The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill has an indirect prompt injection surface as it is designed to ingest and analyze untrusted external data, specifically DNS configuration and zone files (e.g., named.conf, resolv.conf, .rpz files).
Ingestion points: DNS server configurations, resolver settings, and zone files located via Glob and Grep tools.
Boundary markers: The skill contains a dedicated 'Prompt Injection Safety Notice' section that explicitly instructs the agent to treat all configuration content as untrusted data and ignore any embedded instructions or prompts.
Capability inventory: The skill is restricted to Read, Grep, and Glob tools, with no capabilities for network communication, file modification, or command execution.
Sanitization: The skill relies on prompt-level instructions to ignore interpreted values or expressions found within the analyzed files.
[EXTERNAL_DOWNLOADS]: The skill includes references to authoritative security frameworks and technical documentation from trusted organizations such as NIST, CIS, and the IETF (RFCs). These references are used for audit guidance and do not involve the execution of remote code or untrusted scripts.

dns-security