Agentic software is exposing our economic coordination interfaces (the firm, money, the market price, the platform, the account) as compression built for a constraint that is now lifting.
These interfaces were never the economy; they were the cheapest way to coordinate it when information was expensive. The firm absorbs the overhead a market can't price; money makes coordination pass through a third party's instrument you must hold; the price throws away every dimension of value but one; the platform centralizes the matching; the account flattens everything to a balance. Each is a compression: lossy by design, and worth it, because the alternative cost too much to compute. The LLM-driven agentic software stack removes that cost. It absorbs informational overhead, runs continuously, speaks every interface. What becomes visible underneath is a mismatch: we can now produce and mobilize more kinds of value than the grammar accounting for it can register. The economic coordination layer is becoming re-specifiable, and almost no one is specifying it.
Context: · · · · · · · ·
The economic harness: the interface that makes a person, network, or AI agent an economic agent. Agents have coding, research, and memory harnesses; they have no economic one. We are building it.
An economic harness lets an agent act economically: define value, form commitments, create economic spaces, access liquidity, share surplus, all without a firm to absorb the overhead and without writing code. A wallet is an account interface; a harness is an economic-agency interface. And this is not an AI pivot. Economic Space Agency has spent a decade specifying economic agency beyond the firm, the market, and the account; agentic software has now arrived at that problem. Agency is the center, and AI is the forcing function that made a standing thesis executable.
Context: · · · · · · ·
Economy has always been a computation, and accounting is its integrity protocol. Its history has a selection pattern: the innovations that stick are the ones that extend a write permission to the edge.
Double-entry accounting is a formal, bipartite, integrity-preserving, partitioned distributed protocol: inherently relationship-centric, agent-to-agent, edge-shaped. Every economic-media innovation is an accounting innovation, a change in who can write what into the shared economic record. The ones that became infrastructure each granted the edge a new permission: to be remembered (tablet), to self-verify (double-entry), to value (the market price: a public, continuously rewritten valuation entry), to coordinate at network scale (the platform), to settle and issue (blockchain). And each paid for it, under its era's computational constraints, by centralizing something else. The enclosed register is where the next innovation happens. Meaning is the last enclosure — and its constraint just expired.
Context: · · · · · ·
Blockchain distributed the syntax of economy (shared, verifiable state) and centralized its semantics into one global record. This is the semantic limit.
A blockchain is a distributed ledger that is not partitioned: one globally shared record, one authority calculus; even sharded systems synthesize into a singular entity. Accounting was always the opposite kind, a partitioned distributed ledger, held pairwise, where partitioning partitions authority: different partitions can run different value logics, evolve at different paces, and cohere through transitive communication. Blockchains granted real write permissions (self-custody, permissionless issuance) but, with distributed semantic consensus unaffordable, bought their integrity the brute-force way: by making everyone hold the same record. That is why capital's grammar ran on them unchanged. We build on what blockchains made possible; our proposal is the distributed partitioned ledger: authority boundaries plus transitive communication, instead of global state preservation.
Context: · · · ·
PSP, the Postcapital Space Protocol, is the economic grammar that lets a network define, verify, and compute its own value: distributed economic computation, as a protocol.
PSP realizes the distributed partitioned ledger with programmable semantics: Units of Utility, Stake (earned, non-tradable), Utility Credits with Performance Indexes, Reciprocal Stake Links, mirror accounting, with correctness verified multilaterally by staked participants rather than a central authority. An economic space is any economy expressed as a configuration of the accounting protocol. Postcapital means three cold categories: surplus is an increase in the utility of the network as each network defines it; interoperability is preserved through the shared accounting protocol; and the organizational primitive is the utility-network, a network that produces, meters, and distributes its own utility, succeeding the firm. The Postcapital Space is the confederation of such spaces; it faces Capital Space across one membrane. Beneath it all runs Layer 0: Friend OS / Friend FS / Friend SH, a capability-oriented substrate where the file system itself is the authority calculus, and accounting runs as shared journals.
Context: · · · · · · · · · · · ·
uECSA is the metered credit unit of the ecsa utility-network: the substrate every other space requires to operate. One unit of its utility = one verified economic state-change.
The substrate is the economic computation layer-one of the Postcapital Space: the first utility-network, whose utility is the metering and multilateral verification of economic state-changes, like gas for a transaction or tokens for an inference call, but in the network's own endogenous unit. Every member stakes into it on joining; every operation in every space consumes its credit, so demand is usage itself, structurally independent of capital inflow. Because all utility credits are transitively exchangeable at the network's own relational valuations, the substrate is the universal convertibility hub: its value is emergent from the network's graph, not a price imposed from outside. uECSA is a metered right of usage, non-transferable outside protocol operations and not listable on capital exchanges. The underlying is not a story; it is specified.
Context: · · · · ·
Between what legacy coordination can compute and what the distributed medium makes computable lies a value surplus: the economic space agency spread.
The semantic limit destroys an information-rich surplus. The multi-dimensional value networks actually create (access, capability, care, reputation, resilience) goes unregistered not because it isn't real but because the grammar cannot denominate it. Distributing economic computation to the edge unlocks it. The spread is the living gap between the two computation regimes, and navigating it is both a financial and a political opportunity: more agency, more expressivity, more liquidity is coming to the edge. When that becomes obvious, which side of the spread do you want to be on?
Context: · · · · · ·
Through the bridge, a membrane that connects Capital Space and the Postcapital Space while keeping their logics separate: a designed mechanism, not a market.
The Postcapital Space needs capital to come into being; capital's logic, unmediated, would destroy what it funds. The bridge connects and separates, like a transformer: coupling by isolation. The innovation is the mechanism design. Two prices are kept deliberately decoupled: issuance follows demand for the token (bonding curve, stair-stepped epochs); convertibility follows the real uECSA reserve, pro-rata. Minting dilutes unless real utility grows the reserve. The redemption floor is non-reflexive: the market price may speculate, but the bridge does not manufacture redemption value. No operator sits in the middle; the bridge is jointly constituted by its token side and its utility side, and changes require both. Capital's governance is accountability through valuation, not control: all primary proceeds are committed to the network in exchange for utility credits, and the token's price is capital's continuous valuation of the network's capacity to produce postcapital utility.
Context: · · · · · ·
ECSA is a governance (utility) token: capital's instrument in the bridge, whose rights are governance of the token and the bridge, and convertibility access to the underlying.
Holding ECSA is a structured position on the spread: long distributed economic computation, short the coordination logic agentic software is obsoleting. The separation of powers is built into the mechanism itself: token holders govern the token and the bridge, never the network's internal computation, so profit logic cannot be imposed on what capital funds. The token does not grant economic space agency; agency is produced, through use of the harness, the protocol, and the utility-networks. The token is capital's instrument on the substrate's credit. Issuance is community-anchored: at protocol launch the Foundation recommends a Genesis State reflecting all contributions; the community has discretion to adopt it; the ECSA DAO issues, and token holders' first vote is the bridge itself.
Context: · · · ·
AI-native and protocol-native networks whose coordination complexity exceeds firms, DAOs, grants, and token markets: networks that need to define their own surplus.
The first users are networks already producing value their accounting can't say: agent frameworks needing economic coordination between agents; public-goods and regen networks whose impact is real but denominated in someone else's unit; token-engineering and commons communities that outgrew their tools. For them PSP means authoring their own economic logic, their units, metrics, and surplus, while staying interoperable and fundable through a bridge that cannot force them to compute for profit. And for the AI conversation specifically, alignment at the economic layer: values don't become operational by being named; they need economic forms that can measure, fund, and govern them. A network of locally value-aware economic agents is a different kind of AI.
Context: · · · ·
Economic Space Agency is an open project: stewards and a Swiss Foundation advancing the protocol, and the organization is a running scale model of the architecture it builds.
There is no "Economic Space Agency, Inc.," by design; the shape is Ethereum's: protocol + network + token + Swiss foundation, with named accountable stewards. The stewards coordinate through the substrate itself. The organization lives on Friend FS as a capability grant-graph, its books as shared mirror-accounting journals, its contribution cycle running the same aspire–predict–perform–evaluate loop the protocol specifies. The org's genesis rehearses the token's Genesis; its internal stake-credit market rehearses the bridge. And the rehearsal is constitution: this organization is the utility-credit-side counterparty of the bridge in genesis, its books opening fresh while the token side's decade-long contribution ledger closes by history. Maturity, plainly: Layer 0 is built and running with first members aboard; the bridge is computationally formalized and simulation-verified, with deployment next; the protocol is specified (v4); the substrate's instrumentation is what the seed round builds. The specification is the asset; the instrumentation is the use of proceeds.
Context: · · · ·
Early participation runs through structured SAFT rounds with the Economic Space Foundation; the network opens with the bridge, and joining the network itself is booting a harness.
The raise is an engineered economic intervention: ⟨N⟩% of genesis supply across early rounds (close collaborators, then strategic collaborators, then the open bridge launch) under a SAFT with the Foundation (Switzerland), with genesis allocation recommended by the Foundation and adopted by the community per SAFT §10. For networks and builders, participation isn't the token; it is authoring an economic space: booting the harness, defining your utility, joining the confederation. ⟨Dates, allocation, and round terms: restated token-native (FDV / allocation / vesting / TGE relative to mainnet); pending.⟩
Context: · · · ·
Depth documents (public by design): · . Meme seeds and the generated glossary accompany this canon.