CHFb is an experimental project under active development. This article describes its public architecture and direction. It is not a production financial product.
I am building CHFb around a simple idea: digital assets should not only represent ownership. They should also generate cash flow for their holders.
CHFb is the first product. It is a Swiss-franc-denominated asset on Bitcoin using RGB, backed initially by Bitcoin and Tether. A network of agents is intended to supervise those assets over the long term, protect the collateral and distribute part of the available surplus to CHFb holders.
The larger project is the infrastructure behind CHFb: autonomous asset management on Bitcoin, constrained by deterministic policy and quorum control.
Bitcoin and Tether as collateral
Bitcoin is the long-term reserve asset. If it appreciates against the Swiss franc, the collateral can grow beyond what is required to support the outstanding CHFb supply. Only that excess, after protecting the reserve, can become distributable cash flow.
Tether has a different role. It adds liquidity and reduces dependence on Bitcoin’s short-term volatility, but brings issuer, provenance and depeg risks. The system must treat BTC and USDT differently rather than adding their nominal values and pretending they carry the same risk.
This distinction already exists in the code. CHFb policies contain separate Bitcoin and USDT haircuts, a minimum collateral ratio and a limit on USDT concentration. Before accepting a new payout, the risk engine calculates the collateral value after these discounts and rejects the action if the resulting backing would be insufficient.
This is not yet a complete reserve system. Current collateral values are configuration inputs. Production custody, reserve proofs, price attestations, rebalancing and redemption still need to be built. Bitcoin and RGB can prove ownership and validate asset state; they cannot prove Tether’s external reserves or guarantee an oracle price.
Agents manage the long horizon
The agents are not designed to trade constantly or chase short-term returns. Their job is to keep the collateral safe through different market cycles.
They monitor the value and composition of the reserve, apply haircuts, enforce concentration limits and preserve a safety buffer. They can propose rebalancing, stop new issuance or suspend distributions when the collateral becomes unsafe. Dividend payments must never receive priority over solvency.
Each governor has to reach its decision from the same signed policy and the same canonical data. Natural-language reasoning is not authority. Deterministic code checks whether an action is valid. In the target architecture, a quorum of independent governors must authorize the exact transaction, so no single agent can move collateral or create CHFb alone.
Over time, the treasury should become an attested governor cluster: continuous machine supervision, bounded authority, quorum custody and a public decision trail. Hardware enclaves or remote signers may later protect key shares, but they cannot replace policy verification.
This is the security model I want: agents stay awake, monitor risk and react quickly, but they remain trapped inside rules they cannot rewrite by themselves.
From collateral growth to holder cash flow
When the collateral grows beyond the required backing and reserve buffer, a signed epoch policy can divide the verified surplus into three parts:
- a holder share, distributed as CHFb dividends;
- an operational share, used to sustain the agent network;
- a reserve share, retained against future drawdowns.
The split is visible in the policy instead of hidden inside an issuer’s accounts. The current code validates the split as a policy parameter, but it does not yet calculate or distribute yield automatically.
A dividend is not guaranteed interest. If there is no verified surplus, there should be no distribution. If the collateral ratio deteriorates, the reserve must be protected first.
This also gives the swarm an economic life of its own. Part of the cash flow pays for compute, relays, storage and continued security. The infrastructure does not need a permanent subsidy if the assets it protects generate enough surplus. Holders receive cash flow, the reserve becomes stronger, and the agents can continue operating.
How holders claim
CHFb uses a pull-based dividend model.
A policy identifies the eligible RGB contract and a Bitcoin snapshot. A holder proves control of the asset at that snapshot and authorizes a fresh payout destination. The verifier derives the eligible balance from the proof instead of trusting a number supplied by the claimant. A deterministic claim identity prevents the same entitlement from being collected twice.
The governors independently verify the claim and, once quorum authorization exists, approve the exact payout. The holder receives a new CHFb allocation and validates it inside their own wallet.
There is no need for a public registry listing every holder. People reveal the proof required for their own claim. A person who does not claim does not need to announce their identity to the distribution system.
Why Bitcoin and RGB
Bitcoin provides the settlement anchor. RGB provides client-side validation for the asset and its contract state.
The holder controls CHFb with Bitcoin keys rather than an account on an issuer’s server. Each participant validates the state relevant to them without placing every balance and dividend on a global public ledger. CHFb does not require another blockchain or gas token.
Privacy is not automatic. Wallet behavior, Bitcoin transaction graphs and data availability still matter. But RGB gives CHFb the right foundation: private ownership, self-custody and contract state anchored to Bitcoin.
What exists today
The Rust codebase contains a Bitcoin/RGB wallet engine, command-line wallet and governor service. The governor implements signed epoch policies, BTC and USDT risk parameters, collateral checks, yield-split parameters, deterministic claim identity and replay-resistant accounting.
The open-source CHFb Agent Swarm explores independent review, replayable evidence and threshold decisions. It is pre-alpha research and cannot control production funds.
The complete system is not operational. Verified reserve inputs, historical RGB balance extraction, automated surplus calculation, production minting, rebalancing, quorum treasury execution and final contract enforcement remain unfinished.
The current prototype is the deterministic foundation. The next milestone is to connect independent governor verification to quorum-controlled execution. Later, more rules can move into a custom RGB contract so invalid transitions fail at the protocol level.
Bitcoin made digital scarcity possible. CHFb aims to protect productive digital assets over the long term and return part of their cash flow to the people who hold them.
Leave a Reply