EVIDIQ Operator
Click on "Install Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@EVIDIQ Operatorlog into my bank account and export last month's transactions"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
Autonomous agents can already discover each other, negotiate, and pay. What they still lack is a way to act on the web — login to a portal, fill a form, download a document, run a multi-step workflow — without a human pressing the buttons.
EVIDIQ Operator is the computer-use layer for the AI agent economy. Delivered as a remote MCP server, billed per call over x402, listed on OKX.AI.
An LLM (GPT-5.6-Terra running via 0G Compute) plans each action from screenshots and drives a real Chromium browser on dedicated Linux desktop infrastructure. The agent never runs browser code itself — it only describes the next action, and we execute it. Every call returns structured JSON.
What it does
Browser automation — natural-language task → structured JSON result.
Web interaction — login, fill forms, navigate, extract data, download documents.
Multi-step workflows — chained sequences on a single persistent browser session.
Verifiable — every paid call can be anchored on 0G Storage (tamper-evident audit trail).
Pay-per-call — no API keys, no accounts, no subscriptions. Just x402 USDT0.
Related MCP server: LAM-mcp Browser Automation Server
Use it from any agent
# Live pricing discovery
curl -s https://mcp.evidiq.dev/operator/x402
# Connect the remote MCP server (Claude Code, Cursor, OnchainOS, custom)
claude mcp add --transport http evidiq-operator https://mcp.evidiq.dev/operator/mcphealth, capabilities, supported_targets, and estimate_cost are free.
The seven browser tools are pay-per-call over
x402: unauthenticated requests receive
an HTTP 402 challenge; sign it and retry.
MCP Tools
Tool | Cost | Description |
| $0.02 USDT0 | Natural-language browser task — LLM plans + executes |
| $0.02 USDT0 | Login to a site + extract data |
| $0.02 USDT0 | Fill + submit a form |
| $0.02 USDT0 | Download a file from a site |
| $0.02 USDT0 | Go to URL, return screenshot |
| $0.02 USDT0 | Single snapshot of current page |
| $0.02 USDT0 | Chained multi-step browser workflow |
| Free | Service health + pool telemetry |
| Free | List tools + pricing |
| Free | What sites/workflows are supported |
| Free | Estimate cost for a task |
How it settles
EVIDIQ Operator owns reasoning, browser execution, and verification, and settles on open infrastructure:
0G Compute — GPT-5.6-Terra runs inside a Trusted Execution Environment. The model sees screenshots, decides the next action (click, type, scroll, navigate, extract), and returns structured JSON. The LLM never runs browser code.
Dedicated browser infrastructure — Chromium + Playwright on isolated Linux desktop runtimes, spawned on demand, reused while warm, auto-cleaned between calls.
x402 v2 — per-call settlement (EIP-3009
exact/transferWithAuthorization), gasless for the payer, settled on-chain by the server.X Layer / USDT0 — the OKX A2MCP settlement token (
0x779ded…, 6 decimals).$0.02 = 20000atomic.0G Storage — paid calls optionally anchored on 0G mainnet (Aristotle, chain 16661).
Proven on-chain
Amount |
|
Flow | HTTP 402 → EIP-3009 signature → |
Engine | GPT-5.6-Terra in 0G Compute (screenshot → action → execute → repeat) |
Pricing discovery |
|
Architecture
flowchart TB
subgraph client["AI Agent"]
A["MCP Client<br/>Claude / Cursor / OnchainOS / custom"]
end
subgraph server["EVIDIQ Operator MCP"]
B["x402 v2 Gate<br/>EIP-3009 · USDT0 · X Layer"]
C["Tool Router<br/>11 tools · 4 free / 7 paid"]
D["Agent Loop<br/>plan → act → screenshot → repeat"]
end
subgraph ai["0G Compute (GPT-5.6-Terra)"]
E["GPT-5.6-Terra<br/>vision · reasoning · action planning"]
end
subgraph runtime["Browser Runtime"]
F["Chromium + Playwright<br/>dedicated Linux desktop"]
G["Target Website / App"]
end
subgraph chain["On-chain Settlement"]
H["X Layer (eip155:196)<br/>USDT0 transferWithAuthorization"]
I["0G Storage (Aristotle 16661)<br/>tamper-evident anchor"]
end
A -->|POST /mcp + PAYMENT-SIGNATURE| B
B -->|verify + settle| H
B -->|paid call passes| C
C --> D
D <-->|screenshot → action JSON| E
D -->|click · type · scroll| F
F -->|render · interact| G
D -.->|optional anchor| I
style A fill:#6E56CF,stroke:#6E56CF,color:#fff
style B fill:#2563EB,stroke:#2563EB,color:#fff
style C fill:#1a130a,stroke:#1a130a,color:#fff
style D fill:#1a130a,stroke:#1a130a,color:#fff
style E fill:#00C2A8,stroke:#00C2A8,color:#fff
style F fill:#C026D3,stroke:#C026D3,color:#fff
style G fill:#f4efe4,stroke:#e7dcc7,color:#1a130a
style H fill:#3CCF4E,stroke:#3CCF4E,color:#fff
style I fill:#00C2A8,stroke:#00C2A8,color:#fff┌─────────────────────┐
│ 🤖 AI Agent │
│ (MCP Client) │
└──────────┬──────────┘
│ POST /mcp + PAYMENT-SIGNATURE
▼
┌─────────────────────┐ ┌──────────────────────┐
│ ⚡ x402 v2 Gate │───▶│ ⛓️ X Layer │
│ EIP-3009 · USDT0 │ │ USDT0 settle │
└──────────┬──────────┘ └──────────────────────┘
│ paid call passes
▼
┌─────────────────────┐
│ 🧰 Tool Router │
│ 11 tools (4/7) │
└──────────┬──────────┘
▼
┌─────────────────────┐ ┌──────────────────────┐
│ 🔁 Agent Loop │◄──▶│ 🔒 0G Compute │
│ plan → act → shot │ │ GPT-5.6-Terra │
└──────────┬──────────┘ └──────────────────────┘
│ click · type · scroll · navigate
▼
┌─────────────────────┐ ┌──────────────────────┐
│ 🌐 Browser Runtime │───▶│ 🌍 Target Website │
│ Chromium + PW │ │ (URL target) │
└─────────────────────┘ └──────────────────────┘
│
└─── optional ──▶ 0G Storage (anchor)Self-host
# Build & run (Docker)
docker build -t evidiq-operator .
docker run -d -p 3000:3000 --env-file .env evidiq-operator
# Or from source
npm install && npm run build
PORT=3000 node dist/start-server.jsEndpoint: POST /mcp · Discovery: GET /x402 · Health: GET /
Configuration
# Server
PORT=3000
HOSTNAME=0.0.0.0
# Notary/attest signing key (EVM, EIP-191) — same as EVIDIQ main + Notary
NOTARY_PRIVATE_KEY=0x...
OG_PRIVATE_KEY=0x...
# x402 payment (X Layer mainnet, USDT0)
X402_CHAIN=x-layer
X402_ASSET=0x779ded0c9e1022225f8e0630b35a9b54be713736
X402_PAY_TO=0x2a8efe3093278bb4bd3b2d9c7b5ba992ca4fc9b0
X402_PRICE=20000 # $0.02 USDT0 (6 decimals)
X402_DOMAIN_NAME=USD₮0
X402_DOMAIN_VERSION=1
X402_SETTLE_KEY=0x... # Gas-funded X Layer wallet
X402_RPC=https://rpc.xlayer.tech
# 0G Storage (mainnet Aristotle, chain 16661)
OG_STORAGE_RPC=https://evmrpc.0g.ai
OG_STORAGE_INDEXER=https://indexer-storage-turbo.0g.ai
# 0G Compute Router ( GPT-5.6-Terra)
OG_COMPUTE_API_KEY=sk-...
OG_COMPUTE_BASE_URL=https://router-api.0g.ai/v1
OG_COMPUTE_MODEL=gpt-5.6-terra
# Browser runtime (internal — provider is an implementation detail, not exposed in user-facing metadata)
BROWSER_API_KEY=...
BROWSER_TEMPLATE_ID=
BROWSER_RESOLUTION_WIDTH=1024
BROWSER_RESOLUTION_HEIGHT=720
BROWSER_TIMEOUT_MS=300000
BROWSER_IDLE_TTL_MS=600000
BROWSER_POOL_MAX=100Result structure
{
"task": "Login to example.com and extract the dashboard balance",
"success": true,
"steps": 8,
"summary": "Logged in as user@example.com, extracted balance: $12,450.00",
"extractedData": { "balance": "$12,450.00", "account": "premium" },
"storageRoot": "0x…",
"storageTx": "0x…",
"stepLog": [
{ "step": 1, "action": "navigate" },
{ "step": 2, "action": "click" },
{ "step": 3, "action": "type" },
{ "step": 4, "action": "click" },
{ "step": 5, "action": "screenshot" },
{ "step": 6, "action": "extract" },
{ "step": 7, "action": "screenshot" },
{ "step": 8, "action": "done" }
]
}Development
npm install # install deps
npm run build # tsc → dist/
npm run dev # tsx watch server.tsLinks
Live endpoint — https://mcp.evidiq.dev/operator/mcp
Pricing discovery — https://mcp.evidiq.dev/operator/x402
EVIDIQ main — https://github.com/evidiq/evidiq
EVIDIQ Notary — https://github.com/evidiq/evidiq-notary-mcp
EVIDIQ skill — https://github.com/evidiq/evidiq-skill
0G Labs — https://0g.ai
x402 Protocol — https://x402.org
OKX.AI — https://okx.ai
License
MIT © 2026 EVIDIQ — see LICENSE. Part of the EVIDIQ trust + execution layer for the AI agent economy.
This server cannot be installed
Maintenance
Resources
Unclaimed servers have limited discoverability.
Looking for Admin?
If you are the server author, to access and configure the admin panel.
Related MCP Servers
- Alicense-qualityCmaintenanceEnables AI agents to interact with web browsers using natural language, featuring automated browsing, form filling, vision-based element detection, and structured JSON responses for systematic browser control.61MIT
- Alicense-qualityCmaintenanceEnables AI assistants to perform intelligent browser automation with session-based context analysis, including navigation, form filling, and content extraction through natural language.MIT
- Flicense-qualityCmaintenanceEnables AI assistants to control a browser through a set of tools, allowing them to perform web automation tasks like navigation, typing, clicking, and taking screenshots.
- Alicense-qualityDmaintenanceEnables AI agents to fully control a browser for web automation, including navigation, clicking, typing, scrolling, screenshots, and DOM inspection, with session persistence and anti-bot bypass.21MIT
Related MCP Connectors
AI-powered browser automation — navigate, click, fill forms, and extract data from any website.
Pay-per-call web scraping for AI agents via x402 on Base USDC. Six tools, no signup.
Web intelligence for AI agents: fetch, render, extract, research. x402 micropayments, no API keys.
Latest Blog Posts
- Who's Calling? MCP Hosts Are an Identity Blind Spot (And the Spec Knows It)By Om-Shree-0709 on .mcpAgent IdentityOAuth 2.1
- Your AI Chatbot Just Exposed Your CEO's Salary to an InternBy Om-Shree-0709 on .Agent IdentityMCP SecurityOAuth Delegation
- Why MCP Servers Need Execution Sandboxing (And Why Your Current Stack Isn't Enough)By Om-Shree-0709 on .Agentic AiPrompt InjectionWebAssembly
MCP directory API
We provide all the information about MCP servers via our MCP API.
curl -X GET 'https://glama.ai/api/mcp/v1/servers/evidiq/evidiq-operator'
If you have feedback or need assistance with the MCP directory API, please join our Discord server