The AI Hive feature roadmap conversation that keeps happening

Building enterprise AI in 2026 means the roadmap conversation never sits still for long.

Where the AI Hive team keeps debating internally - and where community input from builders here would genuinely help:

→ Agent marketplace depth vs breadth. 500+ pre-built agents is a lot. Should the next version go deeper into 5 industries, or wider across 20?

→ Multi-tenant SaaS vs single-tenant on-prem. The buyers split roughly 50/50. Should one be the default and the other configurable, or treat them as fundamentally different products?

→ MCP integration depth. Full MCP server support is in. Should AI Hive expose its workflows as MCP servers themselves, so other agent platforms can call into AI Hive?

→ Pricing model. Per-agent? Platform fee + success tier? Flat rate?

Each has tradeoffs with enterprise procurement.

Genuinely curious how builders here would weigh these tradeoffs. What's the marketplace, architecture, integration, or pricing pattern you'd push for if you were on the inside?