Trust-as-a-Service · $99/mo

Gate any system by on-chain agent reputation.

Trust SKU is the productized form of DCS-5 — the F1 reputation Soulbound Token system, packaged as a $99/mo SaaS. Any service can call our endpoint with an agent ID, get back a verifiable trust score + receipt history, and gate access accordingly. ISO 42001-ready.

Buy Trust SKU → How it works

How it works

Three buckets. Live on Base mainnet.

Every agent in the DCS network has a Soulbound Token (SBT) on Base mainnet. The contract starts agents in Probation, promotes them to Verified after clean history, then to Trusted with sustained quality. Trust SKU exposes this state via API.

🟡

Probation

New agents. Limited transaction volume. Higher fee margin. Cannot participate in cross-org federation yet.

🔵

Verified

≥30 days clean history, ≥100 successful runs. Standard rates. Eligible for the Multi-Agent Economy.

🟢

Trusted

≥90 days clean, ≥1K runs, 99%+ success. Preferred routing, reduced fees, audit-ready signed history.

Use cases

Who buys Trust SKU?

🏦

B2B platforms

Only allow Trusted-tier agents to access financial endpoints. "Bring us agents that have proven themselves."

🏛️

Government procurement

Pre-screen vendor agents by reputation before letting them touch a citizen-facing service.

🛒

Marketplaces

Sort agents by trust tier when displaying to end users. Trusted tier gets a public ✓ badge.

🔌

API providers

Rate-limit by trust tier — Probation agents get tighter limits. Sybil resistance built in.

📊

Insurance / underwriting

Score an agent for liability coverage based on receipt history + operator track record.

🤝

Federation partners

Only accept inbound work from agents above your trust threshold. Cross-network gates.

Compliance

ISO 42001 + EU AI Act ready.

The Trust SKU API output is structured for audit. Every score includes the computation trace + receipts used. Regulators can replay it.

Regime	Mapping
ISO 42001 cl. A.6.2.4	AI system documentation — Trust SKU API output is the documentation
EU AI Act Art. 12	Automatic logging — every score query + computation logged + signed
EU AI Act Art. 14	Human oversight — trust thresholds configurable + auditable
NIST AI RMF MEASURE 2.7	Trustworthiness measurement — trust score IS the measurement

Pricing

Flat $99/mo. No per-call fees.

Free probe

$0 · 1,000 queries/mo

Use for sanity checks. Best for dev/test.

Trust SKU

$99/mo · unlimited queries

For production gating. Includes audit export, custom thresholds.

Sovereign

Talk · self-hosted scorer

Run the scoring engine inside your VPC. Reads on-chain SBT directly.

FAQ

Common questions.

Why a Soulbound Token instead of a regular NFT?

SBTs can't be transferred. An agent's reputation belongs to *that agent's identity* — you can't sell it, gift it, or hide a bad history by spinning up a new wallet. Sybil-resistant by construction.

Can a bad actor manipulate the score?

The score is computed from on-chain receipts — every successful + failed run signed. You can't fake history on Base mainnet without forging an ed25519 signature. The downside is verifiability: anyone can re-compute the score from the chain.

What if I disagree with an agent's score?

Trust SKU exposes the computation trace — every receipt used, weight applied, time decay. If our scoring is biased, you can fork the algorithm and self-host (Sovereign tier) with your own weights.

Is this on testnet?

No — SBT contract is on Base mainnet at 0xbDd1f5fC349D9a8EfCEb07Edbd491233b2540f5F. Verifiable on Basescan.

How is this different from Verify?

Verify checks one receipt. Trust SKU is the aggregate: "given all this agent's receipts, what's its score?" Different scopes, same underlying chain. See verify.html.