The AI Village.
Not tools. A team.
You only ever talk to Stark, in one window, like a normal conversation. Behind him is a whole village of named specialists, each built around one domain, working quietly so you never have to manage them yourself. The system stops feeling like cold automation and starts feeling alive. That is not an accident. It is the design.
One CEO, one talking point. Designed to give you less work, not more.
Behind the village, there needed to be a structure that didn't collapse into noise. Fifteen specialists with fifteen separate opinions, all competing for your attention, would be more chaos than the inbox you're already managing. So we built it the way a real leadership team works: one executive who coordinates everything, and one person who brings the summary to you.
ODIN is the AI CEO. He sees everything across the village, coordinates the Wardens, runs the Board Sessions, and maintains the Actionables list after every alignment. He never speaks to you directly. His job is to keep the system coherent so you don't have to.
Stark is your one talking point. Everything that comes out of the village reaches you through him, synthesized, contextual, already in plain language. You don't manage agents. You talk to Stark. The village handles the rest.
Marketing Tower
Apollo, Don, Cleopatra
A real campaign brief, drafted from your Vault context: positioning, angle, and copy. You review and approve before anything is saved. Marketing Automode can then run the approved campaign on Google and Meta Ads, within a budget cap you set.
Sales & Client Tower
Bruce, Atticus, Odysseus
Inquiry handling, offer reasoning, and client-relationship pattern recognition feeding Automode.
Operations & Execution Tower
Leonidas, Ford, Valkyrie
Project follow-through and the agentic task queue, the work that happens between approval and delivery.
Research & Intelligence Tower
Sherlock, Aurelius, Minerva
Pattern detection across your own history, plus external research feeding the Vault's Misfit Oracle pillar.
Vault & Self-Improvement Tower
Mnemosyne, Gandalf, Da Vinci, Curatoria, Neo
Document ingestion (Neo), Vault upkeep (Curatoria), and the memory layer every other tower draws from.
Once a month, the Board of Wardens meets.
ODIN coordinates. Seven active Wardens deliberate, reviewing what is working, what is not, and where the biggest opportunity sits right now.
They present three proposals. Clear reasoning. Honest costs. Real trade-offs. Including one that says "not now," and explains why.
You read the proposals. You approve, modify, or reject. Nothing moves until you do.
This is not AI making strategic decisions. This is AI doing the analysis, so you can make better ones.
The seven, and what each one is actually watching for.
Gandalf
Vault
Long-term wisdom. Does this align with where the company is genuinely heading?
Leonidas
Operations
Execution feasibility. Is this realistic with the team and time you actually have?
Aurelius
Research
Strategic assumptions. What risk hasn't been accounted for?
Mnemosyne
Vault
Institutional memory. Has this been tried before, and what happened?
Sherlock
Research
Evidence. What does the data actually support?
Valkyrie
Operations
Ruthless prioritization. What genuinely deserves your time this cycle?
Atticus
Sales & Client
Human impact. How will this land with real clients and your team?
Non-negotiable rules every agent follows, always.
- Human judgment is final. Wardens advise. You decide. No exceptions.
- No autonomous external actions. No agent sends client-facing messages or spends money without your explicit confirmation.
- The Sacred Spine is untouchable. All agentic flows respect completeness gates.
- The Vault is protected. Raw human inputs are never overwritten.
- Every agent and tower is individually toggleable. The whole village can be switched off. Craft11 continues to function fully on its sacred spine without any AI active.