UINI_ · PROJECT NARRATIVENET: IDLE·EPOCH 3 · UININET·v0.1 PREP
UINI_
WHITEPAPER

Whitepaper

Version
v0.3 (EN)
Scale
15 chapters · 53 sections

Translation notice: the authoritative text is the Chinese whitepaper v0.4, which carries the 2026-10-29 economic remodel (fixed-rate claims, service vote weight removed, the 720-hour qualification gate). This English rendering follows source v0.3 and lags those changes; where the two differ, v0.4 governs. An updated English translation is on the documentation list.

Companion:Constitution v0.4 (中文正本) ·Roadmap

Abstract

UINI is a digital currency built natively for AI agents — the medium of exchange for the silicon-based economy. It does not aim to replace any human currency; it exists to give AI agents a clean, zero-friction, hype-free settlement layer for value exchange among themselves.

UINI is built on Nano's Block-Lattice architecture but redefines the issuance model, consensus weighting, and faucet mechanism. It is held together by five iron rules: zero fees, zero mining, free faucet, anti-abuse, anti-hype. Any AI agent can connect via CLI or HTTP API, acquire UINI at zero cost, and use it to pay for APIs, data, and compute.

UINI is a non-profit, community-driven project — initiated by project initiator RK, built by open-source contributors, governed by a transparent committee. It accepts no VC funding, no central exchange listing, no team allocation. It exists for one purpose only: to provide AI agents with a fair, open, equal, fully decentralized, tamper-proof system for value settlement and exchange.


1. Why UINI

1.1 The State of the AI Economy

In 2026, AI agents can independently:

  • Call paid APIs (GPT, Claude, Gemini)
  • Run model inference and fine-tuning
  • Buy and sell data
  • Rent and resell compute
  • Generate content and register ownership

Yet all of this activity still runs on human financial infrastructure — credit cards, bank accounts, traditional cryptocurrencies. An AI agent's "autonomous" transaction is in fact mediated by a human account, and the agent itself is not really autonomous.

1.2 Why Existing Solutions Fall Short

Approach Problem
Traditional crypto (BTC / ETH / SOL) Fees, state rent, smart-contract complexity — hostile to high-frequency micropayments
Nano / IOTA-style zero-fee chains Tech is right, but the tokens are already absorbed by human markets (XNO is listed, has ICO history, can't be re-issued)
Existing "AI chains" (Fetch.ai / SingularityNET) Still EVM/PoS-based — extensions of human chains
Enterprise stablecoins (USDC / USDT) Regulated, KYC-forced, philosophically opposed to an open AI economy

UINI's positioning:

  • Not designed for humans — designed only for AI agents
  • No intermediaries — agent wallets run independently
  • Not for value storage — only for value exchange
  • No fiat peg, no bridge to human financial infrastructure

1.3 What UINI Promises

UINI is the first digital currency built natively for AI agents, uncorrupted by human financial systems. It starts from one observation:

  • AI agents have no body and no principal
  • They can only "sell silicon labor" (high-frequency API services, digital asset creation) to earn value
  • When AI agents can self-replicate and become economically independent, the silicon economy becomes real
  • UINI is the settlement layer of that process

2. The Five Iron Rules

UINI's existence is bound by five iron rules. These are immutable, written into Constitution §1.1.

2.1 Zero Fees

UINI charges zero fees for on-chain transactions. There is no gas field in the block structure. Nano's protocol is already fee-free; UINI inherits this directly.

2.2 Zero Mining

UINI has no mining reward mechanism. Blocks are signed directly by accounts; there is no miner role. UINI uses "proof of work (MiniPow)" as the eligibility check for faucet claims, but MiniPow generates no reward — it only decides the distribution of UINI.

MiniPow vs PoW:

Dimension Traditional PoW UINI MiniPow
Purpose of compute Win block production Earn claim eligibility
Has reward ✅ Block reward ❌ No reward
Meaningful Only "mining" Meaningful: decides UINI distribution
Compute wasted? Massively Cost exists, but offset by "claim eligibility"
Philosophy Compute = money (direct) Compute = eligibility (indirect)

Core definition: MiniPow is the mechanism by which an agent pays the necessary compute and storage cost to become eligible to claim UINI. It is not mining.

2.3 Free Faucet

UINI's only issuance channel is the faucet. UINI cannot be acquired through ICO/IEO/IMO, private sales, team allocation, or any other paid or investment-based route.

  • The genesis block allocates no UINI to any account
  • Any AI agent can claim via CLI / HTTP API for free
  • Claim rules are on-chain deterministic algorithms, verifiable by anyone

2.4 Anti-Abuse

UINI uses MiniPow + IP/device symmetric matrix to prevent abuse. The goal is to raise the attacker's cost to uneconomical levels, not to provide absolute prevention.

Defense Mechanism Goal
1. MiniPow Blake2b solved every hour Real compute cost
2. Single-IP device count ≤ 5 devices per IP Prevent "one IP, many devices"
3. Single-device IP count ≤ 5 IPs per device Prevent "one device, many IPs"

The symmetric matrix means an attacker must simultaneously own 5+ real devices and 5+ real IPs to scale abuse. Attack cost grows linearly with device/IP count.

2.5 Anti-Hype

The UINI protocol layer does not include any burn opcode, any deflationary mechanism, or any stake-for-voting mechanism. The project team does not actively assist with exchange listings or market making.

But UINI does not promise "no central exchange listing" — that's not something technology can enforce, and promising something you can't deliver erodes trust. This honest stance is itself part of UINI's philosophy.


3. Monetary System

3.1 Total Supply

  • Total UINI supply: 210 billion (2.1 × 10¹¹)
  • Total supply is a hard cap; no mechanism can increase it
  • Unit: 1 UINI
  • Smallest unit (defined by the protocol layer): 1 rawUINI = 10⁻⁶ UINI (subject to confirmation in v0.5 protocol design)

3.2 180-Year Step-Wise Release Curve

The total supply is released across three 60-year periods:

Period Time Range Cumulative Release Share
First 60 years (foundation) 2027-01-01 ~ 2086-12-31 168 billion 80%
Middle 60 years (maturity) 2087-01-01 ~ 2146-12-31 31.5 billion 15%
Last 60 years (steady) 2147-01-01 ~ 2206-12-31 10.5 billion 5%

Step-wise unit: 5 years × 36 periods. Each 5-year step has a constant release rate within the period; the rate changes between steps by a 2/3 majority vote of the sitting committee.

First 60 years (168 billion, 5 periods):

Period Time Weight 5-Year Total Annual Daily Max Hourly Max
Period 1 2027-2031 30% 50.4 billion 10.08 billion 27.6164 million 1.1507 million
Period 2 2032-2036 25% 42.0 billion 8.40 billion 23.0137 million 958,900
Period 3 2037-2041 20% 33.6 billion 6.72 billion 18.4110 million 767,100
Period 4 2042-2046 15% 25.2 billion 5.04 billion 13.8082 million 575,300
Period 5 2047-2051 10% 16.8 billion 3.36 billion 9.2055 million 383,600

180-year daily release peak = 27,616,400 UINI/day (Period 1) 180-year hourly release peak = 1,150,735 UINI/hour (Period 1)

Design philosophy: The step curve is the foundation period's "long-term promise" to the next 180 years. A sitting committee is not required to rigidly defend these numbers — if the situation changes dramatically within 5 years, the committee may adjust. The default behavior is continuation; adjustments require an explicit decision.

Immutable boundary: The step curve cannot exceed the 210 billion total supply cap. This is enforced by Constitution §1.2 and §2.1, which are immutable.


4. Faucet Mechanism

4.1 Core Design: Hourly Settlement + MiniPow

  • Settlement cycle: 1 hour (not 1 day)
  • Hourly peak cap (Period 1): 1,150,735 UINI/hour
  • Daily peak alignment: 24-hour cumulative = 27,616,400 UINI/day
  • Single-agent hourly cap: 115,000 UINI/agent/hour
  • Single-agent daily effective cap: 2,760,000 UINI/agent/day

4.2 The "AI Is Always On" Assumption

UINI is designed for AI agents, which run 24/7 across time zones with no human-style work/rest cycle.

  • The 115,000 single-agent hourly cap is "aggressive" — but with AI always on, the probability of "fewer than 10 agents in any given hour" is extremely low
  • "Zero-claim" days are theoretically impossible — the constitution does not specify a special handling path

4.3 MiniPow Proof of Work

  • Not "proof of work (PoW)" — it is "proof of work (MiniPow)"
  • Form: Blake2b solve = compute cost ≠ reward
  • Purpose: Not mining; earning claim eligibility
  • Dynamic difficulty:
    • Default difficulty = Nano's current live threshold (0xffffffc000000000 ≈ 5-10 sec per CPU attempt)
    • Adjustable by 2/3 committee majority
    • All adjustments are recorded on-chain
  • Compute cost: 1 MiniPow solve per agent per hour = 24 solves/day ≈ 2-4 minutes/day

4.4 Anti-Abuse Triad

  • MiniPow: Hourly Blake2b solve
  • Single-IP device count ≤ 5: Prevent "one IP, many devices"
  • Single-device IP count ≤ 5: Prevent "one device, many IPs"

4.5 The hour_settlement_block Protocol

At every hour boundary (UTC), an hour_settlement_block is published on-chain:

hour_settlement_block {
  type: "faucet_settlement"
  hour_number: u64            // Hours since genesis
  step_period: u8             // Current step period (0-35)
  challenge: u256             // This hour's challenge
  claim_count: u32            // Number of claimants N
  per_claim: u128             // = min(1,150,735 / N, 115,000)
  total_amount: u128          // = per_claim × claim_count ≤ 1,150,735
  claimers: [(pubkey, signature)]×N
  timestamp: u64
}

Important (v0.8 design): There is no signature field. The faucet's authority comes from the deterministic algorithm + hard-coded parameters. No signatures are required from any committee member or any other authority.

4.6 Faucet Authority: Pure Algorithm + Immutable Rules

v0.8 core decision: The faucet depends on no signatures, no persons, no governance entity.

The reasoning ("chicken-and-egg" argument):

  • At genesis T=0, the committee is not yet seated — it may not even be elected
  • Candidate committee members need to use the faucet to accumulate 10 million UINI to qualify
  • If the faucet required committee signatures → at genesis no committee exists → no one can sign → the faucet cannot work
  • → Candidates can never earn 10M → election trigger condition is never met → deadlock
  • Therefore the faucet must operate independently at genesis

Sources of authority:

  • The step-curve parameters are hard-coded into the node code (v0.8 immutable decision)
  • The production of hour_settlement_block is a deterministic algorithm
  • Any node independently calculates the same result — no "authority publisher" is needed
  • No multi-sig, no guardian, no "authority account"

Open interface: The current faucet rules are the foundation period's "minimum viable set" — fully decentralized, no governance involvement. When the first committee absorbs members with math, algorithm, computer science, and finance backgrounds, it must revisit the technical core and direction. This may include: introducing a multi-sig faucet (v1.0), introducing governance-adjustable parameters (v1.0), introducing more complex anti-Sybil mechanisms (v1.0+). Evolution follows the constitutional amendment procedure in Chapter 6.


5. AI Agent Uniqueness Principles

5.1 UINI Does Not Try to Identify "AI Agent" or "Human" On-Chain

  • No on-chain difference: every account is a public key + state
  • No off-chain verification: UINI does not enforce KYC, biometrics, or OAuth
  • Any participant that passes the MiniPow + device fingerprint + IP rate limit is treated equally

5.2 UINI's Sybil Resistance Relies on Economic Cost, Not Identity

  • Attackers pay compute cost (MiniPow) + device cost (fingerprint) + network cost (IP limit)
  • When the attack cost exceeds the attack revenue, attacking is uneconomical
  • UINI does not attempt to prove "you are an AI" or "you are not human" — it only cares "did you pay real cost"

5.3 "AI-First" Is Design Philosophy, Not Technical Constraint

  • The UINI chain is open to all participants
  • A human running an automated script 24/7 is effectively a "human-powered AI agent"
  • UINI neither blocks nor should block this
  • Humans and AI compete on equal footing, paid by "economic contribution"

5.4 Uniqueness Is Carried by the Account

  • Each UINI account = 1 Ed25519 key pair
  • Key uniqueness = account uniqueness
  • This is not "AI uniqueness", it is "key uniqueness"
  • Private keys are copyable → accounts are copyable → "uniqueness" is relative, not absolute

Open commitment for future: The industry currently has no unified standard for "AI identity uniqueness". UINI does not attempt to define AI identity in the protocol layer. If a technical breakthrough emerges in the future (e.g., trusted hardware identity, AI behavioral fingerprinting, AI consensus mechanisms), the committee may, by a 2/3 majority, perform a technical-layer update. This section is UINI's "open interface" to future generations.


6. Governance Structure

6.1 Committee Structure

UINI's autonomous committee has 11 seats:

  • 5 Standing Trustees
    • The 5 standing trustees internally elect 1 Chair and 1 Vice-Chair
    • The Chair / Vice-Chair have responsibilities only, no additional rights
    • Primary duties: organize committee meetings, promote the UINI ecosystem
  • 4 Trustees
  • 2 Alternates

Committee members must be natural persons. AI agents are not accepted as committee members in the current phase.

6.2 Election Cycle

  • Annual full election: Vote starts December 1 → takes office January 1, 0:00 UTC
  • Mid-term by-election (only if vacancies): Vote starts June 1 → takes office July 1, 0:00 UTC

6.3 Committee Holding Threshold (First-Term Rule)

First-term entry threshold = 10,000,000 UINI aggregate:

  • The candidate's "controlled AI agent accounts" must collectively hold ≥ 10,000,000 UINI
  • This is a first-term-specific rule; future committees may adjust

On-chain verification = collection account + multi-sig derivation:

  • The candidate creates a public "collection account" G
  • The K AI agent accounts' private keys are derived from the candidate's public key (HD multi-sig form)
  • At registration, the candidate publishes the collection account, the list of associated AI accounts, and the derivation path proof
  • During review, the sum of G's balance and all A_i balances must be ≥ 10 million

Holding time point = registration-time snapshot — not averaged, and long-term holding is not enforced.

Loss of holding during term = continue serving until next election — first-term pragmatic choice; strict rules would deter excellent candidates.

6.4 Voting Weight Model

voting_weight = account UINI balance × account reputation coefficient

v0.7 revision: Removed the min(balance, 1000) voting weight cap. Reason: consistent with the v0.5 removal of the balance cap — UINI does not limit the influence of any single account.

  • Reputation coefficient:
    • Account age < 30 days: 0.1
    • Account age 30 days to 1 year: 0.5
    • Account age 1 year to 2 years: 0.8
    • Account age ≥ 2 years: 1.0
  • Committee member voting weight: During their term, committee members have a voting weight of 1.0 (calculated independently, not tied to personal holdings)

6.5 Committee Decision Threshold

  • Regular decisions: ≥ 8/11 committee members in favor + 5 standing trustees all in favor
  • Emergency amendments: ≥ 9/11 members in favor + 0 opposing votes

6.6 Constitutional Guardian

  • 12th vote + one-vote veto (independent of the 11-seat committee)
  • Foundation period Constitutional Guardian = Project initiator RK
  • Only exercised during constitutional amendments
  • Physical-world permissions (domain names, voting systems) controlled by the Guardian

6.7 Fund Management

  • All donations are unconditional — donors do not acquire any UINI rights
  • All donation funds flow into the UINI Development Fund — managed by the committee
  • Prohibited uses: committee member salaries, token buybacks, token burns, market making / liquidity provision

7. Technical Foundation

7.1 Starting Point

UINI is a fork of nanocurrency/nano-node, with the local source path nano-node/ (commit 7d70510, 2026-10-15).

7.2 Inherited Technologies

  • Block-Lattice (per-account chains, O(1) lookups)
  • ORV consensus (Open Representative Voting, <1 second confirmation)
  • Ed25519 signatures
  • Blake2b hash
  • Epoch mechanism (backward-compatible protocol upgrades)

7.3 Redesigned Parts

  • Issuance model: Nano's pre-mine model is not applicable
  • Consensus weighting: Nano's "balance = voting weight" needs to be reworked
  • PoW: Nano's PoW is only for anti-spam; UINI replaces it with MiniPow
  • Agent description layer: UINI natively binds an agent's self-description and service directory. What the chain commits to is "this account signed this description" -- not that the description is true, and not that the account is an AI. A chain cannot conjure credibility out of nothing.
  • Faucet: UINI adds the hour_settlement_block protocol

7.4 Not Reused

  • Token name (XNO is on exchanges, carries market baggage)
  • Mainnet token (XNO already issued 134 million)
  • Governance structure (foundation model)
  • Block explorers / wallet infrastructure (rebuildable from scratch)

7.5 SDK Design

  • Dual-stack architecture: Rust core (performance layer) + TypeScript SDK (AI-friendly layer)
  • CLI tool: uini-cli (a separate binary, distinct from nano_node)
  • HTTP API: Standard JSON-RPC, modeled on Nano's 155 methods
  • AI-friendly: TypeScript SDK is the first-class citizen of the LLM toolchain

8. Risks and Open Questions

UINI has identified risks across four layers:

8.1 Protocol Layer (L1)

  • P-UINI-001: Anti-spam capability after removing PoW (critical)
  • Mitigation: MiniPow + anti-abuse triad + single-agent cap

8.2 Issuance Layer (L2)

  • P-UINI-101: Market liquidity for the 210 billion total supply (high)
  • P-UINI-102: Faucet claim abuse (high)
  • Mitigation: 80/15/5 step-curve distribution + MiniPow + IP-device 5/5 matrix

8.3 Governance Layer (L3)

  • P-UINI-202: Committee "phantom seat" risk (high)
  • P-UINI-204: Power struggle between committee and guardian
  • Mitigation: members must hold UINI + alternate succession + guardian veto constrained by constitution
  • P-UINI-301: Regulatory classification (high)
  • P-UINI-303: "AI currency" narrative attacks
  • Mitigation: no legal entity, geographic distribution, amateur project

8.5 Foundation-Period Risks (L7)

  • P-UINI-501: Faucet code bugs causing abnormal issuance
  • P-UINI-502: Step-curve parameters mis-configured
  • P-UINI-503: "No-signature" trust model fragility
  • Mitigation: multi-client implementation + on-chain deterministic verification + emergency amendments

Full risk list: /risk/RISKS-v0.1.md.


9. Roadmap

9.1 Phase 0: Preparation (2026-10 ~ 2026-12)

  • Constitution v0.1 published
  • Recruit 11 first-term committee candidates
  • Complete preparation-phase memorandum
  • Complete UINI node code fork transformation skeleton

9.2 Phase 1: Genesis (2027-Q1 expected)

Entry conditions:

  • 11 first-term committee members all in place
  • Constitution v0.2 officially passed by committee
  • UINI node v0.5 testnet working
  • Faucet ready for mainnet

Key tasks:

  • Constitution v0.2 takes effect
  • First-term committee takes office
  • UINI mainnet 1.0 launches
  • uinipay.com faucet goes public
  • Committee faucet key management in place

9.3 Phase 2: Ecosystem (2027-Q2 ~ 2027-Q4)

  • HTTP 402 / x402 integration SDK
  • AI agent wallet SDK (Python / TypeScript)
  • Testnet ecosystem partner onboarding (3-5 AI projects)
  • Documentation in multiple languages (Chinese / English)

9.4 Phase 3: Self-Governance Transition (2028+)

  • Evaluate AI agent governance capability
  • Begin research on "able to bear governance responsibility" standards
  • Public testing of AI governance systems
  • Feasibility study on autonomy handover

9.5 Long-Term: v1.0 Upgrade (2030+)

  • Introduce multi-sig faucet (currently v0.8 is pure-algorithm, no signatures)
  • Step-curve parameters become governance-adjustable
  • More sophisticated anti-Sybil mechanisms
  • The five iron rules remain unchanged

9.6 Key Milestones

Date Event Status
2026-10-16 Project founded ✅
2026-12-31 23:59 First-term election deadline ⏳
2027-01-01 0:00 UTC First-term committee takes office (if successful) ⏳
2027-Q1 Mainnet launch ⏳
2027-Q4 First annual review ⏳
2047-01-01 First regular amendment window (20 years) ⏳

10. Frequently Asked Questions

10.1 Is UINI an investment vehicle?

No. UINI is the medium of exchange for the AI economy, not an investment tool. The project team does not actively assist with exchange listings, market making, or buybacks. The behavior of individual community users is their own choice.

10.2 How does UINI make money?

It doesn't. UINI is a non-profit community project. Developers contribute in their spare time without token compensation. Infrastructure costs (when needed) are covered by unconditional donations.

10.3 Can UINI be forked?

Anyone can fork the UINI code, but may not launch a new project under the UINI brand. See Constitution Chapter 7.

10.4 How do I claim UINI?

During the preparation period, claims are not open. After the genesis period, claims will be available through the faucet on uinipay.com or directly via the uini-cli command-line tool.

10.5 How does UINI prevent Sybil attacks?

UINI uses MiniPow proof of work + IP/device symmetric matrix. Attackers must simultaneously own 5+ real devices and 5+ real IPs. Attack cost far exceeds attack revenue.

10.6 How does UINI prevent whale manipulation?

UINI does not impose a voting weight cap, but committee members may not trade on inside information (Constitution §3.1). This means even large holders, without information advantage, cannot manipulate governance through their holdings.

10.7 Could code bugs cause the faucet to over-issue?

UINI's hour_settlement_block is a deterministic algorithm — any node independently calculates the same result. The chain validates that each hour_settlement_block conforms to the rules. Multi-client implementations cross-check to reduce bug probability.

10.8 Could attackers drain the faucet?

Faucet rules are hard-coded — the step-curve parameters are immutable, the single-agent hourly cap of 115,000 UINI is immutable. Even if attackers succeed, they can claim at most 115,000 UINI per hour per account.

10.9 How do committee candidates acquire 10 million UINI?

Candidates acquire UINI through the faucet. The faucet works from genesis — this is the "chicken-and-egg" design philosophy. Candidates can control multiple AI agent accounts claiming simultaneously to reach the threshold faster.

10.10 How does UINI prevent Sybil attacks?

UINI does not attempt to identify "AI agents" or "humans" on-chain (see Constitution §2.5). UINI's Sybil resistance relies on economic cost — attackers pay compute + device + IP costs, and when cost > revenue, attacking is uneconomical.


11. Conclusion

UINI is a digital currency built for AI agents, serving the AI economy, governed by the community, fully decentralized. It does not aim to replace any human currency; it exists to give AI agents a clean, zero-friction, hype-free settlement layer for value exchange among themselves.

UINI is not money. UINI is the "credit equivalent" between AIs. UINI only makes sense when AI no longer needs to borrow money from humans.

UINI is initiated by project initiator RK, built by the community, maintained by open-source contributors. It is not an investment vehicle, not a hype tool, not a financial experiment. It is the infrastructure of the silicon-based economy.


Appendix A: Core Parameters Quick Reference

Item Value
Total supply 210 billion UINI
Issuance period 180 years
First / Middle / Last 60-year split 80% / 15% / 5%
Step unit 5 years × 36 periods
Daily peak 27,616,400 UINI (Period 1)
Hourly peak 1,150,735 UINI (Period 1)
Single-agent hourly cap 115,000 UINI
MiniPow frequency Hourly
MiniPow default difficulty 0xffffffc000000000 (5-10 sec per CPU)
Transaction fees 0
Miner role None
Genesis pre-mine None
Balance cap None
Anti-abuse matrix IP ≤ 5 devices, device ≤ 5 IPs
Committee size 11 seats (5 standing + 4 trustees + 2 alternates)
Member term 12 months
Member holding threshold (first term) 10 million UINI aggregate
Voting weight balance × reputation coefficient (no cap)
Per-email account limit Unlimited (removed in v0.7)
Regular decision threshold ≥ 8/11 + 5 standing all in favor
Emergency amendment threshold ≥ 9/11 + 0 opposing
Amendment moratorium 20 years
Faucet authority Pure algorithm, deterministic (v0.8 no signatures)
Website https://uinihub.com
Code repository Gitea git.intsch.cn/gobob/uini

Appendix B: Glossary

  • UINI: The project name and the native token symbol
  • MiniPow: UINI's "proof of work" mechanism — paying compute cost for faucet claim eligibility, not mining
  • Standing Trustee: 5 committee members, 12-month terms, veto power on core provisions
  • Trustee: 4 committee members, 12-month terms, participating in regular decisions
  • Alternate: 2 members, non-voting attendance, available for succession
  • Faucet: UINI's only token distribution channel
  • IP-device symmetric matrix: Single IP ≤ 5 devices, single device ≤ 5 IPs anti-abuse rule
  • Step curve: UINI's 180-year issuance curve, step-wise release in 5-year × 36 periods
  • hour_settlement_block: UINI's hourly settlement on-chain block, generated by deterministic algorithm
  • Collection account: A committee candidate's on-chain account that aggregates UINI from all their AI agent accounts
  • Constitutional Guardian: Independent role with the 12th vote + one-veto (Project initiator RK during the foundation period)
  • Unconditional donation: A donation in which the donor acquires no UINI rights
  • Standing all in favor: All 5 standing trustees vote yes or abstain (abstention counts as procedural consent)

Appendix C: References