Governance
Governed by people. Supported by technology.
Comutability is community-led and AI-assisted. AI helps with research, drafting, code and testing. Authority and responsibility for decisions stay with people you can name.
Where this stands: one document is approved, the Founding Charter. The rest on this page is proposed and can still change.
What is approved, and what is not
The founder approved the Founding Charter on 28 October 2026. Putting a document in a repository does not approve it: approval is a decision a named person records.
- Status: Approved
Founding Charter v1.1
What Comutability is for, and the ten principles it works by.
- Status: Draft, not ratified
How Secure Net is governed
The proposal process for the protocol, who decides, and what AI may not do. In the repository as a draft.
- Status: Being drafted, not yet published
ORG-001: Community-led organisation, human-controlled operations
Separates community governance, day-to-day operations and AI-assisted work, with people responsible for each.
It will be published for comment before any decision.
- Status: Being drafted, not yet published
GOV-002: Independent protocol stewardship
Keeps the protocol independent of any one application, keeps the applications separate, and keeps people in charge of AI.
It will be published for comment before any decision.
- Status: Draft, not ratified
Governance proposal
The first proposal for how community decisions are made.
- Status: Draft, not ratified
AI development policy
What AI tools may do in our work, and what needs a person.
- Status: Draft, not ratified
Operating handbook
Day-to-day working practices.
Who would decide what
Proposed, not adopted. For now the founder coordinates direction and infrastructure.
- Mission and the charter
- The founding steward for now, moving to a wider body as the community grows
- Changes to the protocol standard
- The Technical Steering Committee
- Each application's features
- That application's maintainers
- Security-sensitive changes
- A qualified human reviewer, as well as the maintainers
- Releases and deployments
- An authorised human maintainer
- Drafts, code suggestions and routine testing
- People, assisted by AI
What AI may not do on its own
These limits come from the draft documents above. We work by them already, and say so when we have fallen short.
Where we have fallen short
Parts of Secure Net and Commonveil were written and put in place by an AI assistant at the founder's request, before these rules were drafted and without independent review. They are recorded as awaiting human review, not as approved.
- Vote, or sit on a committee.
- Approve its own security-sensitive changes.
- Merge protected changes or publish a release without a person authorising it.
- Change governance rules or the charter.
- Give itself more access.
- Spend money, or appoint or remove people.
From one steward to many voices
We are starting with founder stewardship, because someone has to be accountable while there is not yet a community to be accountable to. The intention is to move towards elected representation as a genuine contributor community forms.
The proposals set out routes for contributors to question a decision, to have it reviewed by someone who did not make it, to remove representatives who are not doing the job, and to propose changes in leadership. None of that is in force yet. How people are chosen, how votes are counted and how appeals work each need their own proposal.
Funding does not buy a say. Sponsors and donors would get no control over community decisions or the protocol standard by paying.