Appearance
System Map
What we're building, how the pieces connect, and what's real vs. simulated.
This document is for builders. The brief answers why this should exist. The whitepaper explains how the mechanism works. This answers: what are we building and where is each piece right now?
The Loop
A project launches a token. Trading generates fees. Those fees fund governance. Governance produces decisions. Good decisions attract more trading. More trading funds better governance.
trading volume
→ fees → emissions pool grows
→ bigger bounties → more deliberation participation
→ better decisions → higher confidence signal
→ more trading volumeThe claim — and what the simulation has to prove — is that this loop is self-sustaining: governance generates more value than it costs to run.
The Full Sequence
Where Each Piece Lives
| Component | Fast Sim | Real Agents | Not Yet Built |
|---|---|---|---|
| Market AMM / bonding curve | ✓ /emitter/runs | — | on-chain contracts |
| Fee split (40/50/10) | ✓ | — | on-chain contracts |
| Emissions pool compounding | ✓ | — | on-chain contracts |
| Customer / Provable Work | ✓ | — | SDK integrations |
| Qualitative Work submission | ✓ /runs | — | — |
| Proposal voting / threshold | ✓ /runs | — | — |
| Anode AMM (pricing, pools) | ✓ /runs | — | — |
| Deliberation economics (speak/vote/settle) | ✓ /runs | — | — |
| Claim content + reasoning quality | bots (stochastic) | to build | — |
| Facilitator (turn management) | — | ✓ /deliberate | — |
| Adversarial analysis | ✓ /runs | — | — |
| Emitter ↔ Governance bridge | planned | — | — |
| Execution (decision → action) | — | — | to build |
| Quality signal → market feedback | planned | — | — |
Two Simulation Layers
Layer 1 — Fast Sim Stochastic bots, Monte Carlo over parameter space. Runs in the browser in seconds, costs nothing. Answers: does the mechanism math work? Used for parameter exploration and adversarial analysis. Lives in /runs (governance) and /emitter/runs (launchpad).
Layer 2 — Real Agents Actual Claude instances with isolated context, economic stakes, and MCP tools to interact with the simulation state. Runs slowly and costs money. Answers: does the mechanism change how agents actually reason? The interesting finding is the gap between what Layer 1 predicted and what Layer 2 actually produced.
The Facilitator (/deliberate) is the coordination layer for Layer 2 — it manages turn-taking, extracts claims, tracks votes, calls settlement. It's a referee, not a player.
The Key Number
Self-funding ratio = additional trading volume attributable to governance / total governance cost
> 1.0— governance pays for itself. The loop is self-sustaining.< 1.0— governance is subsidized. The loop needs external funding to run.
Everything in the simulation is infrastructure to produce this number under realistic conditions.
What Gets Updated Here
This document is a living reference. When a component moves from "to build" to "built," update the table. When the sequence changes, update the diagram. When the self-funding thesis changes, update the key number section.
The brief and whitepaper are versioned and archived. This document just reflects current reality.