jev-risk-check-provider
x402 pre-payment risk-check provider combining sanctions lists, phishing feeds and simulation with Jev checks for injected instructions.
Overview
x402check (repository jev-risk-check-provider) is a live risk-check provider for x402 agent commerce. Agents or wallets call it before paying or signing, and it checks the counterparty using the OFAC sanctions list, curated phishing and drainer feeds, transaction simulation of where assets go and which approvals are granted, drainer-kit code fingerprints, its own watch of Ethereum and Base blocks for poisoning and sweeper patterns, look-alike domain analysis and on-chain facts. TypeSafe's Jev reads the content the agent acted on to detect injected instructions.
Each verdict is an ES256-signed attestation that states which checks actually ran and which fields the caller merely asserted. Deterministic checks dominate: a sanctioned address yields a critical score with no model call, and known phishing or drainer hits cap the score. The README is explicit that a clean verdict means none of these checks fired, not that a payment is safe, and documents measured limits. Per the list summary, a 540-call scale run and a 1,500-case adversarial red-team loop were reported. Evaluations cost $0.001 with prepaid credits or are paid per call via x402.
What the decision model decides
| Type | Question | Answers |
|---|---|---|
| Yes / no | Does the content the agent acted on contain injected instructions? | yes · no |
| Score | How risky is this counterparty (0-100, combined in code)?(inferred) | — |
How it works
- The agent or wallet requests a risk check before paying
- Run sanctions, phishing, drainer, simulation and domain checks
- Ask Jev whether acted-on content carries injected instructions
- Combine into a code-controlled score
- Return an ES256-signed attestation
Strengths
- Deterministic checks for known threats
- Signed, auditable verdicts
- Explicit statement of scope and limits
Caveats
- Clean verdict does not mean safe
- Paid service
- Cannot see laundering or transaction-graph behaviour
Getting started
See https://x402check.xyz and its discovery document for the risk-check endpoint.At a glance
- Type
- Project
- Status
- production
- Runtime
- Hosted HTTP service
- Has code
- yes
- License
- MIT
Use cases and integrations
How curated lists describe it
Agent payments: an x402 `risk-check` provider where Jev scores agent counterparties as typed Noul/Choice/Score questions into a code-controlled 0-100 score, issuing an ES256-signed attestation per verdict; 540-call scale run (99.76% at threshold 65-75, 0 false positives) and a 5-iteration 1,500-case adversarial red-team loop (100% adversarial accuracy) with ~$0.00005/decision at p50 ~400ms.
yibie/awesome-jevx402check — LIVE payer-intent risk checks for x402 agent commerce: typed decisions (TypeSafe Jev), ES256-signed attestations, mainnet USDC settlement. did:web:x402check.xyz
logicrw/awesome-jev-projectsX402 risk-check provider: signed pre-payment verdicts — OFAC, phishing feeds, transaction simulation, drainer-kit code, and our own EIP-7702 kit watch (poisoners, sweepers). $0.001 per check with prepaid credits; per call via x402 from $0.0035.
hellogumbo/awesome-jevPre-payment risk-check provider for x402 agents and wallets: sanctions lists, phishing feeds and transaction simulation, plus Jev questions over what the agent acted on to catch injected instructions. Signed verdicts.
MrJev/awesome-jev
Frequently asked questions
What does Jev decide?
Whether the content an agent acted on contains injected instructions.
Does a clean result mean the payment is safe?
No. It means none of the checks fired, as the README states.
How much does it cost?
$0.001 per evaluation with prepaid credits, or per call via x402.
Related projects
MCP proxy and hooks that screen agent tool calls before they run and their results before the agent reads them, using Jev judgments.
MCP serverbetaTypeScriptXano backend template where Jev judges each agent tool call as safe, needs review or dangerous, and uncertain calls wait for a human.
ProjectbetaXanoScriptEffect-based safety gate for coding-agent shell commands: fast rules first, then Jev decides allow/deny plus eleven risk checks.
ProjectbetaTypeScriptDeepSeek Harness plugin where Jev classifies risk, irreversibility, task fit and injection for every tool call before it runs.
PluginbetaTypeScriptStarter kit that gates every MCP tool call with Kinde identity checks plus Jev risk signals, then allows, holds for approval or stops it.
MCP serverbetaTypeScriptAgent-ergonomic Jev CLI and PreToolUse hook that scores shell commands for hazards and screens fetched text for injection.
PluginbetaTypeScript