Open Questions

These are the questions that should be answered through use, not over-designed upfront.

Workspace and Bot Design

1. Which MyOS-Core function should be the first full execution-team pilot: Competitive Intelligence, Content Engine, or another internal function?
2. Should dedicated Telegram bots be created per active workspace, or should Hermes profiles and Telegram topics be tested first?
3. What is the minimum workspace template that Arijit will actually use without friction?

ELT Design

4. Which ELT roles are always useful for idea review, and which should be summoned only when relevant?
5. Should ELT critique be a visible artifact every time, or only for larger decisions?
6. What is the threshold for Athena to challenge Arijit versus proceed with assumptions?

Execution Team Design

7. Which execution roles need skills first?
8. Should project workspaces use .company/ JSON state immediately, or only after repeated manual coordination pain?
9. What quality gate is mandatory before Athena says a deliverable is done?

Knowledge and Memory

10. Which notes should Hermes load by default for My OS?
11. Which notes should remain vault-only and never enter default runtime context?
12. How should raw Telegram transcripts be summarized and retired?

Current Recommendation

Pilot this model on one real workspace before creating more machinery.

Best candidates:

• Competitive Intelligence: already has a ledger, daily/weekly cadence, and clear consumption path.
• PRISM is no longer a MyOS-Core workspace candidate. It is an external system with its own Hermes and webapp.