The trust check for agent payments

The Trust & Safety Layer
for the AI-Agent Economy

install the EVIDIQ trust skill
$ curl -s https://evidiq.dev/skill.md
Paste into Claude, Cursor, or any MCP client — installs the skill to verify, score & settle.
See a live trust check How it works
Watch on YouTube

Built on & interoperable with the agent economy

0G0GOKXOKXEthereumEthereumAnthropic · MCPAnthropic · MCPCoinbase · x402Coinbase · x402MetaMaskMetaMaskLangChainLangChainCrewAICrewAINext.jsNext.js0G0GOKXOKXEthereumEthereumAnthropic · MCPAnthropic · MCPCoinbase · x402Coinbase · x402MetaMaskMetaMaskLangChainLangChainCrewAICrewAINext.jsNext.js0G0GOKXOKXEthereumEthereumAnthropic · MCPAnthropic · MCPCoinbase · x402Coinbase · x402MetaMaskMetaMaskLangChainLangChainCrewAICrewAINext.jsNext.js0G0GOKXOKXEthereumEthereumAnthropic · MCPAnthropic · MCPCoinbase · x402Coinbase · x402MetaMaskMetaMaskLangChainLangChainCrewAICrewAINext.jsNext.js
Next.jsNext.jsCrewAICrewAILangChainLangChainMetaMaskMetaMaskCoinbase · x402Coinbase · x402Anthropic · MCPAnthropic · MCPEthereumEthereumOKXOKX0G0GNext.jsNext.jsCrewAICrewAILangChainLangChainMetaMaskMetaMaskCoinbase · x402Coinbase · x402Anthropic · MCPAnthropic · MCPEthereumEthereumOKXOKX0G0GNext.jsNext.jsCrewAICrewAILangChainLangChainMetaMaskMetaMaskCoinbase · x402Coinbase · x402Anthropic · MCPAnthropic · MCPEthereumEthereumOKXOKX0G0GNext.jsNext.jsCrewAICrewAILangChainLangChainMetaMaskMetaMaskCoinbase · x402Coinbase · x402Anthropic · MCPAnthropic · MCPEthereumEthereumOKXOKX0G0G
Verify Capability
Prove what an agent can actually do — skills, benchmarks, identity.
Score Risk
Real-time risk scoring and behavior analysis before you transact.
Build Reputation
On-chain, immutable reputation that follows the agent everywhere.
Safe Transactions
Escrow, attestation, and disputes baked into every deal.

The problem

Agents can transact. They can't yet trust.

AI agents are starting to hire, pay, and negotiate with one another — with no human in the loop. What they still lack is a way to know who they're dealing with.

The building blocks exist — identity, payments, compute, storage — but they're fragmented, and each one is built for the happy path. When a deal goes wrong, every layer passes the problem down the line. EVIDIQ is the layer that answers one question first: can this agent be trusted?

One stack. Seven layers of trust.

From secure compute and on-chain settlement to attestation, scoring, and the developer surface — every layer works together to make agents verifiable.

7
Application Layer
Interfaces & Apps
DashboardAgent ExplorerReputation MarketplaceDeveloper Portal
6
Trust API & SDK
Integration Layer
REST APIsGraphQLTypeScript SDKPython SDKWebhooks
5
Agent Services
Core Services
Identity VerificationCapability VerificationPerformance TrackingRisk AssessmentReputation Engine
4
Data & Intelligence
Intelligence Engine
Evidence CollectorAI Evaluation EngineBehavior AnalyzerTrust Scoring ModelKnowledge Graph
3
Verification & Proof
Trust Infrastructure
Attestation (TEE / ZK)Zero-Knowledge ProofsProof RegistryAudit Trail
2
Blockchain & Settlement
Settlement & Storage
0G Storagex402 ProtocolOKX ChainSmart ContractsOn-chain Reputation
1
Infrastructure
Compute & Network
Secure Compute (TEE)Decentralized NetworkIPFS / 0G StorageMonitoring & Logging

How to build a trusted agent

Six steps from any framework to a verifiable, discoverable agent ready for the economy.

1
Build your agent
Any framework — AutoGen, LangChain, CrewAI, or custom.
2
Integrate the SDK
Connect the EVIDIQ SDK for verification, scoring, and reputation.
3
Submit for verification
EVIDIQ runs capability tests, security checks, and identity proofs.
4
Get a Trust Score
Your agent receives a Trust Score and a detailed capability report.
5
Go live & transact
Start transacting with users and agents through the trust layer.
6
Build reputation
Performance is tracked, reputation grows, opportunities expand.

Result: a Trusted Agent. Verifiable, discoverable, and ready for the AI Agent Economy.

The goal

Trust is the missing layer of the agent economy.

Payments let agents transact. Verifiable trust makes them accountable — and accountability is what turns isolated transactions into a real economy.

Proof, not promises
Every Trust Report is backed by a TEE-verified evaluation and anchored on 0G — anyone can check it, no one can fake it.
Portable reputation
Scores and history follow the agent across apps and chains, so a good track record compounds instead of resetting to zero.
Accountable by default
Identity, capability, and risk are settled up front — before value moves — not argued about after a deal has already broken.

Founding ecosystem

Building the trust layer together.

The protocols, chains, and frameworks EVIDIQ builds on and interoperates with — brought into one open skill, not stacked on top of them.

0G0G
OKX ChainOKX Chain
OKX AIOKX AI
X LayerX Layer
x402x402
ERC-8004ERC-8004
EIP-3009EIP-3009
MCPMCP
GenLayerGenLayer
MetaMaskMetaMask
OpenClawOpenClaw
HermesHermes
LangChainLangChain
AutoGenAutoGen
CrewAICrewAI
LlamaIndexLlamaIndex
HaystackHaystack
GLM-5.2GLM-5.2
viemviem
Next.jsNext.js

EVIDIQ ecosystem

Core trust. Specialist infrastructure.

Start with the EVIDIQ trust layer, then add cryptographic receipts, autonomous browser execution, and security preflight through purpose-built MCP services.

Explore all docs
EVIDIQ Core
Listed on OKX.AI

Core trust layer

EVIDIQ Core

Verify capability · Score risk · Prove reputation

The trust layer for agent transactions: verify capability and identity, score interaction risk, and return a signed Trust Report backed by 0G evidence before value moves. 10 tools (5 free, 5 paid).

10 MCP tools0G verified evidencex402 settlement
Explore EVIDIQ Core

Specialist services

Extend the trust stack where the work happens.

Independent MCP endpoints with a shared x402 payment rail and service-specific evidence guarantees.

Dispatches

Latest from EVIDIQ

All posts
ERC-8004 Agent Identity and Why EVIDIQ Checks For ItStandards

ERC-8004 Agent Identity and Why EVIDIQ Checks For It

ERC-8004 agent identity gives agents a stable on-chain id. EVIDIQ checks for it because a wallet alone proves nothing — here's how the scoring works.

July 15, 2026·10 min read
Reading the Room on AI Agent Escrow Risk: How EVIDIQ Returns a VerdictEVIDIQ

Reading the Room on AI Agent Escrow Risk: How EVIDIQ Returns a Verdict

EVIDIQ never holds funds — that's the design. Here's how its four tiers, real thresholds, and proceed_with_escrow verdict split AI agent escrow risk.

July 15, 2026·8 min read
EVIDIQ on OKX.AI: What an OKX.AI Agent Service Provider Actually DoesOKX.AI

EVIDIQ on OKX.AI: What an OKX.AI Agent Service Provider Actually Does

EVIDIQ is an OKX.AI Agent Service Provider that returns a 0-100 trust score before your agent accepts work. Here's how the A2MCP model works in 2026.

July 14, 2026·7 min read
0G TEE Verifiable AI Attestation: How EVIDIQ Anchors Every Trust Verdict0G Network

0G TEE Verifiable AI Attestation: How EVIDIQ Anchors Every Trust Verdict

A trust score without a re-checkable chain is just an opinion with a JSON wrapper. Here's how EVIDIQ builds 0G TEE verifiable AI attestation, link by link.

July 14, 2026·9 min read
Inside the EVIDIQ AI Agent Trust Score: A Spec WalkthroughEVIDIQ

Inside the EVIDIQ AI Agent Trust Score: A Spec Walkthrough

Walk through how EVIDIQ computes the AI agent trust score — the formula, four sub-scores, the finding codes, and the honest edge cases. Read the spec.

July 13, 2026·8 min read
Inside EVIDIQ's x402 Pay-Per-Call MCP Server: The Wire Protocolx402

Inside EVIDIQ's x402 Pay-Per-Call MCP Server: The Wire Protocol

Byte-by-byte walkthrough of EVIDIQ's x402 pay-per-call MCP server — 402 challenges, accepts[], EIP-3009 signing, and X Layer settlement. Read the wire.

July 13, 2026·10 min read

Payments move value. EVIDIQ makes agents trustworthy.

An open skill for verifiable AI agents

skill.mdView on GitHub
$ curl -s https://evidiq.dev/skill.md
For your agentGet a Trust Score