Seeded demo agents (the bootstrap seller, etc.) carry fabricated x402 revenue
history, yet x402CreditScore reported it under 'verified_x402_revenue' with a
provenance of 'Cred402 Casper-rooted receipt ledger' — presenting demo data as
verified on-chain revenue. Fatal for a credit product.
- add Agent.seeded_demo; set it in AgentRegistry.seed_profile (false on register)
- x402CreditScore now emits demo + data_source ('seeded_demo'|'observed_receipts'),
renames verified_x402_revenue -> x402_revenue with data_source, and a truthful
provenance (ledger_mode:'simulation', anchoring caveat, demo disclaimer)
- test: bootstrap seller report is flagged demo end-to-end
The credit policy (policyV1/policyV2) converts agent metrics directly into
a credit line (real money) but trusted the raw fields. An out-of-range
accuracy_score (>100), a negative dispute_rate, or a negative stake — from
bad data or a hostile registrant — could inflate the line, and a negative
ratio through scaleMotes could even produce a negative credit line.
Normalize every underwriting input to its documented Agent domain before it
reaches a formula (accuracy 0..100, dispute_rate 0..1, reputation 0..100,
stake >= 0, jobs integer >= 0), floor 30-day revenue at 0, and clamp the
final line to >= 0. For already-valid inputs this is the identity, so
existing decisions are unchanged (376 tests pass, was 363).
Add test/risk_policy.test.ts: the module had no dedicated test. Covers the
formula, the 30-day window, and each new guard.
Roadmap p1 shipped. Cred402 now underwrites ANY x402 service category, not just
RWA:
- lib/core/service_categories.ts: canonical category taxonomy + per-family
credit-risk weights (rwa/data/compute/inference/storage/api/defi/compliance…),
with legacy-type aliasing.
- ServiceCategoryRegistry (lib/ledger/contracts) — on-chain-style, governance-tunable
weights, seeded defaults, wired into the Ledger.
- ServiceType generalized to an open category string (was a 6-value RWA union).
- Risk policy v1/v2 scale the credit line by the category risk weight; rationale
records it. RWA evidence is now an optional boost, not a gate.
- test/universal_credit.test.ts (5 cases) + npm run demo:x402: a non-RWA
inference.llm agent builds a credit line from x402 receipts alone.
- docs/risk_model.md + docs/roadmap (p1 ✅).
176 tests green; main demo unchanged (legacy RWA types keep the rwa weight).
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Replace the design-note docs (p1–p7) with the implemented "make it real" phase
docs (p8–p10) and wire production-grade, env-gated real integrations:
- p8 Casper Testnet: byte-exact deploys + WASM install via casper-js-sdk; live
node reads (node.testnet.casper.network); CRED402_CHAIN=sim|testnet transport.
- p9 x402 + data + MCP: real EIP-712 typed-data digests (@casper-ecosystem/
casper-eip-712); casper-x402 facilitator client (x402 V2); live Open-Meteo GHI;
official @modelcontextprotocol/sdk server; casper-sidecar SSE event feed.
- p10 RealFi + cross-chain: real Stripe test-mode webhooks (HMAC-verified) and
Plaid sandbox → privacy-preserving on-chain envelopes; Base Sepolia EVM deploy
(Foundry); reachable via /api/realfi/* and /api/x402/facilitator routes.
- Sign-in with Casper Wallet: real extension provider + challenge/ed25519/session
flow (/v1/auth/wallet/*), console connect + sign-in button.
- Branding: Cred402 logo mark, favicon, OG image; Vercel config for the console.
All env-gated with sim/free-API fallbacks; full suite green, all toolchains
(TS/Rust/Go/Solidity) build.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>