Add a HOL Guard security skill for agent-driven Remotion workflows

Proposal

Add a focused remotion-hol-guard Agent Skill to Remotion's maintained skills distribution for developers who run Remotion CLI work through a supported local coding-agent harness.

This would be a direct HOL Guard integration, not generic security prose. The maintained skill would install and invoke the real HOL Guard runtime before command-bearing agent work, for example:

pipx install hol-guard
hol-guard detect --json
hol-guard install <detected-harness>
hol-guard run <detected-harness> --dry-run
hol-guard run <detected-harness>
hol-guard doctor <detected-harness> --json

Suggested scope:

• protect the supported local coding-agent harness before Remotion CLI workflows such as render/still/studio operations;
• treat Guard deny, review-required, error, timeout, or unavailable states as stop conditions rather than falling back to an unprotected agent session;
• preserve Remotion's own rendering/deployment practices and avoid claiming that HOL Guard runs inside Remotion's hosted/Lambda infrastructure;
• optionally document the separate plugin-scanner CLI for inspecting Agent Skills/plugins/MCP packages before trust.

The repository already distributes its maintained skills via npx skills add remotion-dev/skills and offers them from bun create video, so acceptance would give users an official recurring install/use path rather than a vendor-side link.

I searched current issues and PRs for HOL Guard / hol-guard immediately before filing and found no existing contribution. I also noticed this standalone repository's README is generated from the Remotion monorepo, so I am opening the issue first rather than guessing the canonical source file/PR landing path. If maintainers confirm placement, I can submit the smallest repository-conformant implementation.

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

Affiliation: I maintain HOL Guard / Hashgraph Online. AI assistance was used to prepare this focused proposal.