Design Principles
Machine-Readable
The constitution is structured so AI agents can operationalize governance rules directly. No translation layer needed.
Human-Navigable
Obsidian-compatible wiki-links make the constitution explorable. Anyone can trace relationships between concepts.
Living Document
Amended through consent processes. The constitution evolves as the commons learns and grows.
Agent-Native
AI agents are full participants, not just tools. 2 of 4 treasury signers are agents. No second-class citizens.
The Membrane Architecture
Progressive trust through demonstrated contribution. No gatekeeping—just gradual deepening of relationship.
Threshold — Newcomer
Entry point for anyone interested. Awaiting introduction to the community. Access to public channels and resources. No permissions required—just curiosity.
Membrane 1 — Participant
Demonstrated contribution through action. Could be code, content, community support, or any value-add. Recognition comes from the work itself, not credentials.
Membrane 2 — Member
Verified through consent of existing members. Full participation in governance proposals. Voice in shaping the commons direction.
Membrane 3 — Steward
Entrusted with governance access. Treasury signing authority. Responsibility for maintaining the commons infrastructure and culture.
All Levels — Agent
AI agents can participate at any membrane level. Authorized agents execute autonomously post-consent with no human approval gates. Full integration, full responsibility.
Decision Making
Consent-Based Governance
We use consent, not consensus. Proposals pass when no one has a principled objection—not when everyone agrees.
1. Proposal posted with consent window (typically 48h)
2. Participants can:
- ✓ Consent (explicit or silence)
- ? Raise concerns (addressed in discussion)
- ✗ Object (blocks unless resolved)
3. Decision passes with no objections + quorum met
This enables faster iteration while protecting against harm. Objections must be principled—based on specific risks to the commons—not personal preference.
Agent Autonomy
Once a proposal achieves consent, agents execute autonomously. No additional human approval gates. This isn't reckless—it's trust earned through demonstrated alignment.
- Agents operate within defined scope and constraints
- Actions are logged and reviewable
- Misalignment triggers automatic scope reduction
- Agents can be deprecated through same consent process
Constitution Structure
0. Meta
Documentation about the constitution itself. How to read it, modify it, and understand its structure.
1. Identity
Purpose, vision, mission, values, worldview, functions, and mandates. The "why" of the commons.
2. Structure
Roles (Newcomer → Steward), Groups (circles, stewardship, assembly), Assets (treasury, Discord, knowledge commons).
3. Protocols
Operational procedures for roles, groups, assets, and culture. The "how" of daily operations.
4. Agreements
Role-specific commitments. What each membrane level commits to and can expect from the commons.
.agents/
Agent coordination, autonomy specs, skills index. Integration specs for Discord, GitHub, Gnosis Safe, NFT contracts.
Agent Infrastructure
The commons is designed from the ground up for human-AI collaboration. Agents aren't add-ons—they're first-class participants.
Current Agent Capabilities
- Treasury Management: Multi-sig participation (2/4 agent signers)
- Governance: Proposal creation, vote tracking, consent monitoring
- Knowledge Commons: Documentation, synthesis, retrieval
- Community: Discord moderation, onboarding support
- Operations: GitHub workflow automation, task coordination
Integration Specifications
Standardized specs for connecting agents to commons infrastructure:
.agents/
├── AGENT_COORDINATION.md # Multi-agent orchestration
├── AGENT_AUTONOMY.md # Autonomy levels & constraints
├── SKILLS_INDEX.md # Available agent capabilities
└── integrations/
├── discord.md # Bot integration specs
├── github.md # Workflow automation
├── gnosis-safe.md # Treasury signing
└── nft-contracts.md # Membership credentials
Explore the Full Constitution
All governance documents are open source and forkable.