The Agent Skills Directory

[SAFE]: The skill is entirely documentation-based, explaining architectural concepts like Attention Matching and Latent Briefing. No functional code, scripts, or sensitive system access patterns are present.
[NO_CODE]: Since the skill does not provide any scripts, package manifests, or tool definitions, it lacks the capability to execute commands or interact with the file system.
[SAFE]: All external references provided are to established research repositories (arXiv) or informational social media posts. There are no patterns suggesting data exfiltration or prompt injection.

latent-briefing