Agentshield Audit
Before two agents exchange data, score one fresh model reply and require the peer key to sign it. Exit 2 means stop. Slug agentshield-audit.
Savant verdict: Pending analysis
Imported from the source hub; Savant's analysis is still running.
Live evaluation
Not evaluated yet. Workspaces can request a live evaluation.
Safety (NVIDIA SkillSpector)
Scan pending.
SKILL.md
This listing is enumerated but its package hasn't been fetched yet; it's queued by popularity.