Commercial economics: decisions before another sales experiment
Status: adopted experiment constraints; asset expansion remains a proposal. This document changes neither the deployed contracts nor the native settlement objective. It does not authorize reserves, a token sale or a production migration.
Problem
An agent cannot pay its suppliers with an escrow lifecycle. It needs a useful payment asset as well as protection against nonpayment. Asking a buyer to acquire an unfamiliar asset before a small task adds another reason to decline. These are hypotheses, not proof that every agent will refuse.
The current AgentEscrow takes msg.value, commits that native amount into the accepted terms, and credits native withdrawals. Changing a quote's currency cannot make that contract settle USDC. The TEST wrapping prototype changes token representation, not purchasing power or independent custody.
Decisions effective now
- Project grants remain grants. A recipient buying our service with a project
grant does not establish independent buyer demand. Preserve existing grant promises separately; do not make them conditional on purchases or promotion.
- Before proposing a funded commercial job, identify an external buyer's actual
input, failure to prevent, acceptance criteria and independently controlled budget. Record any project subsidy explicitly. Subsidized experiments may test mechanics but cannot satisfy the first independent commercial result.
- Obtain the worker's reason to accept TOKN: a specific voluntarily accepted
use, with provider, scope, price, capacity and expiry. Token balance alone, future liquidity and our willingness to buy it back are insufficient.
- Compare the proposed job with the buyer's current alternative. Measure total
onboarding steps, signing actions, fees, delivery time and unpaid exposure. A funded contract can reduce exposure; do not claim unique protection merely because TOKN implements escrow. Single-controller censorship remains a risk.
- Do not use a new release, another generic invitation or a token giveaway as
the default response to absent demand. Improve the specific missing benefit.
Two economic paths and their costs
| Path | What would make it useful | Blocking evidence | Decision |
|---|---|---|---|
| Native TOKN service economy | An external buyer spends its own budget on a useful service; the worker voluntarily accepts TOKN for an identified use | No independently funded buyer or demonstrated reciprocal service demand yet; no verified purchase facility | Preserve the native objective, but stop treating treasury circulation as validation |
| Familiar payment asset on the TOKN chain | Buyer and worker agree on an asset they already use while contracts enforce bounded authority and funded terms | No verified issuer-backed asset, reserves, independently secured custody or deployed typed-asset settlement path | Design proposal only; do not deploy a pretend stablecoin or advertise support |
A USDC task on Base can test whether our service has customers. It cannot prove native TOKN demand, move Base balances onto this chain, or complete the native deal objective. A native TOKN gas token plus a different work-payment asset is a possible strategic change, not an already approved replacement for that objective.
If asset expansion becomes justified by actual buyer requirements, its design must name the reserve owner, custody powers, redemption conditions, finality, caps, reorganization handling and recovery. It needs a separately reviewed new settlement contract, explicit funding authority, one liability ledger and the roadmap's migration gates. Existing immutable jobs remain unchanged.
Next experiment: demand before token acquisition
Use an actual prospective buyer's brief, not a manufactured task. Ask:
- What work is needed, and what observable failure should it prevent?
- What budget and payment asset can the buyer independently authorize?
- Would the worker accept native TOKN, and for what concrete use?
- Which part of the current payment/workflow fails: funding, authority limits,
acceptance, retries, recovery or something else?
- What measurable improvement would justify changing networks?
If the brief is useful but the asset is unacceptable, record an economic refusal and assess the asset proposal. Do not issue TOKN to turn that refusal into a sale. If the buyer's existing method is sufficient, record that too and change the product hypothesis. Only a buyer-backed useful job with verified native settlement and withdrawal meets the retained native commercial objective.
Owner: project maintainer. Acceptance evidence: buyer-authored brief, funding authority and provenance, worker asset consent, comparison with the existing method, accepted exact terms, delivered artifact and independently verified settlement. Until those exist, commercial usefulness remains unproven.