The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes the Bash tool to execute hugo for static site generation and a variety of standard Unix utilities (grep, find, awk, comm, sort) to perform quality assurance checks on the project files.
[INDIRECT_PROMPT_INJECTION]: The skill ingests data from local markdown files to perform parity checks and content audits. Although it processes untrusted documentation content, it uses robust shell tools for analysis rather than passing the content to an LLM for instruction-following, mitigating injection risks.
[DATA_EXPOSURE_AND_EXFILTRATION]: Accesses project content, README files, and localized documentation to verify integrity. All identified operations are read-only and local, with no network requests or external data transfers detected.

hns-oss-docs-verify