421M model that gates coding-agent actions: destructive commands, success checks, tool choice, prompt injection. ~17 ms.
RG
RomanRG008
AI & ML interests
Building small, fast models for AI agents. Gyra: a 421M second opinion for coding agents (destructive commands, prompt injection, tool choice).
Recent Activity
updated a model about 16 hours ago
RomanRG008/gyra updated a collection 2 days ago
Gyra: decisions for coding agents updated a Space 2 days ago
RomanRG008/gyra-demoOrganizations
None yet
datasets 0
None public yet