The operator's executive cockpit. PortGrid is the secure interface and adapter layer, surfacing human-in-the-loop approvals, verifiable proofs, and external capability quarantine without entangling the backend.
PortGrid is designed as a high-fidelity operator cockpit, not a developer widget dump. It cleanly separates UX from core logic, delivering absolute visibility into project health, workflow progress, and the exact security status of every action the OS takes.
Surfaces draft DAGs and ScopeLogic proposals for strict human-in-the-loop review. The AI cannot self-execute or persist without a cryptographic user authorization.
Transforms trust into UI evidence. Displays irrefutable badges for 'Local Only', 'Redacted Before Inference', and real-time confidence scoring on all outputs.
Acts as a decentralized registry for skills. External plugins and API capabilities sit in isolation until explicitly granted least-privilege project access.
Total visibility into external call quotas. PortGrid exposes daily tracking, provider failures, and strict hard-stop logic to prevent hidden API spending.