Operations

Claude / Anthropic

API key readiness; AI agents disabled until guardrails ship.

Claude / Anthropic

Default Model
claude-sonnet-4-5
AI Runtime
disabled
Outbound AI
requires human confirmation

AI agents remain disabled until guardrails ship. AI must never mark a communication log as approved or sent.

Anthropic · Required for go-live

Anthropic API key present in environment

Completed

ANTHROPIC_API_KEY set in the eapinder-ops Vercel project for Production + Preview. AI agents themselves remain gated by ANTHROPIC_AI_ENABLED and human-confirmation guardrails (tracked separately).