Proposal: add a HOL Guard security companion skill for agent-driven Cloudflare workflows

I'd like to contribute a focused `hol-guard` security companion skill to Cloudflare's official agent-skills collection.

The goal is to make HOL Guard itself available in the same maintained skill surface developers already use for agent-driven Cloudflare work, rather than asking Cloudflare to add a new generic security API or hook.

Proposed scope:

- add a named HOL Guard skill in the repository's preferred skill location;
- install and invoke the actual HOL Guard tooling from the skill, using only currently supported Guard workflows;
- cover practical agent-security tasks such as scanning agent/plugin artifacts, checking Guard status/policy/evidence, and applying Guard around supported local command/tool workflows before high-impact operations;
- keep Cloudflare-specific examples factual and limited to behavior HOL Guard can actually enforce today;
- avoid any claim that HOL Guard runs inside Workers or transparently intercepts Cloudflare's hosted runtime.

Acceptance would be easy to validate: a user installs the skill through Cloudflare's existing distribution path, follows the skill, and reaches the real `hol-guard` / scanner tooling rather than a provider-neutral checklist.

HOL Guard: https://github.com/hashgraph-online/hol-guard
Canonical maintained skill source: https://github.com/hashgraph-online/hol-guard-plugin/tree/main/skills/hol-guard

If this placement fits the collection, I can submit the smallest focused PR following the repository's preferred structure.

Start by inspecting the repository's preferred skill location and existing distribution path, then compare the canonical maintained source at skills/hol-guard. Verify that the installed skill reaches the real hol-guard and scanner tooling and covers only supported Guard workflows; done when the stated installation and acceptance flow works.

Written by the indexing model from the issue text.