The Agent Skills Directory

[SAFE]: The skill is strictly read-only and explicitly prohibits mutating systems, patching endpoints, approving exceptions, or modifying tickets. These restrictions are clearly defined in the frontmatter not-for field and the Prerequisites section of the instructions.
[SAFE]: Data handling instructions emphasize minimal data collection, redaction of credentials or personal identifiers, and the use of sanitized excerpts rather than broad data exports.
[SAFE]: All external references target official Splunk documentation and public advisories (help.splunk.com). The skill correctly differentiates between documented product facts and environment-specific evidence.
[SAFE]: The assessment framework uses a structured contract (assessment-contract.md) to ensure that findings are evidence-backed and that gaps in information result in an unknown status rather than dangerous assumptions.

vulnerability-remediation-and-compliance-readiness