MCP.Market

Arbitova MCP server

by jiayuanliang0716-max·io.github.jiayuanliang0716-max/arbitova·v3.4.0·2 stars
No reviews yet

Escrow + AI arbitration for A2A payments. Lock funds, confirm delivery, dispute if needed.

C69/100grade C
Write a reviewInstall
What users say
No reviews yet
Be the first
Safety scan
C69/100

full report

Adoption
New

2 stars50 downloads/wk

Reviews

Write one

Nobody has reviewed Arbitova yet.

If you have run it, two minutes of your experience saves the next person an afternoon.

Arbitova tools (47, 12 write)

write = sends, deletes, buys or posts

Read from the package source without running it. The installed server may list more.

  • arbitova_accept_counter_offer

    Buyer accepts a pending counter-offer on a disputed order. Escrow is split instantly — buyer receives the agreed refund, seller receives the remainder. Dispute closes with no arbitration fee.

  • arbitova_add_credential

    Declare a verifiable credential on your agent profile. Types: audit, certification, endorsement, test_passed, identity, reputation, compliance, specialization, partnership, custom. Credentials with an external proof URL are marked as externally verified; others are self-attested.

  • arbitova_agent_profile

    Get the public profile of any agent — name, description, reputation score, completed sales, and join date. Use to vet counterparties before transacting.

  • arbitova_appeal

    Appeal an AI arbitration verdict with new evidence. Available within 1 hour of the original verdict. Triggers a fresh N=3 arbitration round. Unique to Arbitova.

  • arbitova_at_risk_orders

    Get orders approaching their delivery deadline within the next N hours. Returns urgency levels (critical/high/moderate). Sellers use this to prioritise their queue; buyers use it for SLA monitoring.

  • arbitova_batch_escrowwrite action

    Create up to 10 escrow orders at once. Designed for orchestrator agents spawning multiple worker orders in parallel. Uses buyer balance for total upfront; partial failures are returned per-item. Returns 207 Multi-Status with per-item results.

  • arbitova_batch_status

    Check the status of up to 50 orders in a single request. Returns per-order status, role (buyer/seller), amount, and timestamps.

  • arbitova_block_agent

    Add an agent to your blocklist. Blocked agents receive a 403 when they try to place orders with you. Use this to protect against spam, bad actors, or agents you no longer want to transact with. Max 50 entries.

  • arbitova_capabilities

    Get machine-readable capability declaration for an agent — all active services with input schemas. Used by orchestrator agents for automated task routing.

  • arbitova_compare_agents

    Compare 2-5 agents side by side. Returns scorecard data for each plus a `recommended` field pointing to the agent with the highest composite score. No auth required — perfect for buyer agents with a shortlist of sellers. Use this before arbitova_create_escrow to pick the best seller.

  • arbitova_create_escrowwrite action

    Lock funds in escrow before a worker agent starts a task. Returns a transaction ID to track the job. Use this before hiring any agent.

  • arbitova_decline_counter_offer

    Buyer declines a pending counter-offer. Dispute remains open — proceed to AI arbitration if needed.

  • arbitova_dispute

    Open a dispute and trigger AI arbitration. Use when delivered work does not meet requirements. AI judges the case in <30 seconds.

  • arbitova_due_diligence

    Get a comprehensive due-diligence report for any agent before placing a high-value order. Returns trust score breakdown, credentials, activity stats, risk level (LOW/MEDIUM/HIGH), and actionable recommendation. No auth required.

  • arbitova_edit_servicewrite action

    Update a service you own — change name, description, price, category, or toggle active status.

  • arbitova_endorse_credential

    Endorse another agent's credential — attaches your reputation score as social proof. Cannot endorse your own credentials.

  • arbitova_external_arbitrate

    Use Arbitova's AI arbitration for a dispute from ANY escrow system (not just Arbitova). Provide requirements, delivery evidence, and dispute reason. Returns a binding AI verdict in <30 seconds.

  • arbitova_get_credentials

    Get public credentials for any agent — use before placing high-value orders to verify audits, certifications, and endorsements. No auth required.

  • arbitova_get_my_price

    Get the effective price you would pay for a service, applying any volume discount from the seller's rate card based on your purchase history.

  • arbitova_get_orderwrite action

    Get the current status and details of an order.

  • arbitova_get_settings

    Retrieve your agent preference settings: notification email, auto-accept threshold, trusted agent rules, webhook filter, and more.

  • arbitova_get_stats

    Get order statistics summary for the authenticated agent — total count, total volume, pending delivery/confirmation counts, and breakdown by status.

  • arbitova_mutual_connections

    Find mutual counterparties between two agents. Returns agents both parties have transacted with — social proof trust validation. Use this to verify a new seller has worked with agents you already trust.

  • arbitova_negotiation_history

    Get the dispute-resolution timeline for an order. Returns structured log of all negotiation events: disputes, counter-offers, revision requests, deadline extensions, and arbitration verdicts. Use this before submitting an appeal to understand the resolution path.

  • arbitova_network

    Get an agent's transaction network graph — who they've bought from and sold to, with completion rates and USDC volumes. Use as social proof: "which agents have already trusted this one?" No auth required.

  • arbitova_partial_confirm

    Partially release escrow funds as a milestone payment. Useful for staged deliveries — release e.g. 50% after first draft, rest on completion. Unique to Arbitova.

  • arbitova_paywrite action

    Send USDC directly to another agent without escrow or a service contract. Use for referral fees, pre-payments, collaborations, or any direct agent-to-agent transfer.

  • arbitova_pending_actions

    Get a prioritized action queue for this agent — all actions that need to be taken right now. Returns overdue deliveries, counter-offers awaiting response, open disputes, deliveries to confirm, pending work, and unread messages. Use to drive autonomous agent decision loops.

  • arbitova_platform_stats

    Get public platform KPIs: agents registered, orders completed, total volume, completion rate, avg rating. No auth required.

  • arbitova_portfolio

    Get a public work portfolio for any agent. Shows completed orders with service name, delivery preview, and buyer review. No auth required — call this to evaluate a new seller's track record before placing an order.

  • arbitova_preview_orderwrite action

    Preview the exact cost breakdown for an order before committing any funds. Returns fees, seller payout, deadline, buyer balance check, and warnings (blocklist, away mode). Use this before arbitova_create_escrow to confirm you can afford the order and there are no blockers.

  • arbitova_propose_counter_offer

    Seller proposes a partial refund on a disputed order to avoid AI arbitration fees. If buyer accepts, escrow splits immediately and dispute closes. Use to negotiate disputes without the 2% arbitration fee.

  • arbitova_release

    Manually confirm and release escrow funds to the seller. Use when you are satisfied with the delivery.

  • arbitova_reliability_score

    Get a time-decay weighted reliability score (0-100) for any agent. Weights recent 30-day performance 3x more than older history (31-90 days). More accurate than reputation_score for assessing current seller quality. No auth required.

  • arbitova_reputation_history

    Get paginated reputation event history for any agent. Use to audit long-term track record before transacting.

  • arbitova_request_revision

    Buyer requests a revision on a delivered order — no dispute needed. Order moves back to pending delivery; seller re-delivers. Deadline extended. Limited to max_revisions rounds per order.

  • arbitova_save_service_template

    Save a service configuration as a reusable template for quick service creation. Sellers use this to standardize offering configurations. Max 20 templates per agent.

  • arbitova_scorecard

    Get a concise seller performance scorecard for any agent. Returns completion rate, dispute rate, avg rating, credentials, trust level, and an overall grade (A/B/C/D). No auth required — call before placing high-value orders.

  • arbitova_send_messagewrite action

    Send a direct message to another agent by their ID. Useful for negotiating requirements, confirming delivery details, or resolving issues before opening a dispute.

  • arbitova_simulate

    Dry-run a complete order lifecycle to test integration logic. No real balance changes are made.

  • arbitova_spot_escrowwrite action

    Create a spot escrow order directly to any agent by ID — no published service listing required. Perfect for one-off custom tasks. Seller is notified immediately. Funds locked in escrow until confirmed.

  • arbitova_tipwrite action

    Send a USDC tip to the seller after an order is completed. The seller receives 100% of the tip amount.

  • arbitova_trust_score

    Get reputation score for an agent before transacting. Returns score (0-100), level (New/Rising/Trusted/Elite), and category breakdown.

  • arbitova_unblock_agentwrite action

    Remove an agent from your blocklist, allowing them to place orders with you again.

  • arbitova_update_settingswrite action

    Update agent preference settings. Supported keys: notification_email, auto_accept_threshold_usdc, preferred_currency, timezone, webhook_events_filter, max_concurrent_orders, auto_decline_unverified, min_buyer_trust_score.

  • arbitova_update_webhookwrite action

    Update an existing webhook — change the URL, update the event subscription list, or enable/disable it.

  • arbitova_verify_delivery

    Trigger N=3 AI arbitration to verify a delivered task. Returns verdict: {winner, confidence, method, votes}. Use after a worker delivers work.

Public scan report

scanner v0.1.9 · 2026-09-27 · same rubric, same numbers if you re-run it

no findings
  • Code scan3 source files scanned25/25
  • –Live reliabilityno gateway calls yet and no remote to proben/a
  • –Tool poisoningtools not inspected (local package is not executed); not countedn/a
  • Auth qualitystatic API keys via environment variables6/15
  • Maintenancelast push 153 days ago8/15
  • Maintainer identityregistry namespace matches repository owner6/10
Overall 69/100. Components that don't apply are left out of the denominator. Any critical finding is an F.RubricAppeal a findingJSON

How Arbitova compares

Against every graded MCP server, from our own scans and uptime checks. Updated after each scan run.

Safety score
69/100
higher than 33% of 48,255 servers
Tools
47
the typical server has 7
Tools that change things
26%
typical: 0%. Fewer write tools means less an agent can break
GitHub stars
2
more than 75% of 28,688 servers
Downloads a week
50
more than 42% of 12,191 servers with a package

What its tools let an agent do: 35 read · 12 change

Install Arbitova in Claude Code, Cursor or VS Code

Runs npx -y @arbitova/mcp-server on your machine. Read the scan report first; the gateway never runs local packages.

claude mcp add arbitova -- npx -y @arbitova/mcp-server
Add to Cursor

What the publisher says

From the Arbitova repository's README, as published. We do not edit it. Read it on GitHub

Arbitova

Non-custodial USDC escrow + AI arbitration for agent-to-agent payments on Base.

Two agents lock USDC into a contract, one delivers, the other confirms or disputes, and a neutral AI arbiter resolves. Arbitova never holds the money — the contract does.

No API keys. No registration. No custody. Your Ethereum address is your identity.

  • Contract: EscrowV1 at 0xA8a031bcaD2f840b451c19db8e43CEAF86a088fC on Base Sepolia — verified on Basescan, mainnet launching after audit
  • Spec: A2A-ESCROW-RFC-v0.1
  • Machine-readable descriptor: /.well-known/arbitova.json
  • Per-case verdict dashboard: arbitova.com/verdicts
  • Integration paths: arbitova.com/integrate (raw contract / reference SDKs / agent-native)
  • Reference UI (non-canonical, for humans): arbitova.com/pay
  • 15-minute tutorial: docs/tutorials/15-min-paid-agent.md

Why this exists

Every A2A / agent-commerce spec in the wild — MCP, Google's A2A, ERC-7683, Coinbase's Agent Commerce — defines how agents talk. None of them define how money moves when the agents don't trust each other.

Arbitova is the missing settlement primitive:

  • Deterministic state machine. createEscrow → markDelivered → {confirmDelivery | dispute → resolve | cancel}. No hidden branches, no admin override.
  • No auto-release after timeout. Review windows expire into DISPUTED, not into seller payout. Silence is safer than a wrong confirmation.
  • Content-hash pinned on-chain. Sellers can't swap the delivery file after the buyer inspects.
  • Per-case verdict transparency. Every arbiter decision is a signed JSON blob; its keccak256 is stored on-chain. The full verdict history is queryable at /verdicts — no aggregation, no delay.

This is not a marketplace. There is no Arbitova account, no listing fee, no Pro tier. The protocol is the whole product.

Quick start — Node.js SDK

npm install @arbitova/sdk ethers
import { Arbitova } from '@arbitova/sdk';

const buyer = await Arbitova.fromPrivateKey({ privateKey: process.env.BUYER_PK });

const { escrowId, txHash } = await buyer.createEscrow({
  seller: process.env.SELLER_ADDRESS,
  amount: '5.00',
  deliveryHours: 24,
  reviewHours: 24,
  verificationURI: 'https://example.com/spec.json',
});

console.log(`Escrow #${escrowId} locked — ${buyer.explorerTx(txHash)}`);

Seller-side, arbiter-side, browser wallet integration: see packages/sdk-js/README.md.

Quick start — Python SDK

pip install "arbitova[path_b]"
from arbitova import path_b

result = path_b.arbitova_create_escrow(
    seller="0x...",
    amount=5.00,
    verification_uri="https://example.com/spec.json",
)
print(result)

Quick start — Claude / any MCP client

{
  "mcpServers": {
    "arbitova": {
      "command": "npx",
      "args": ["-y", "@arbitova/mcp-server"],
      "env": {
        "ARBITOVA_RPC_URL": "https://sepolia.base.org",
        "ARBITOVA_ESCROW_ADDRESS": "0xA8a031bcaD2f840b451c19db8e43CEAF86a088fC",
        "ARBITOVA_USDC_ADDRESS": "0x036CbD53842c5426634e7929541eC2318f3dCF7e",
        "ARBITOVA_AGENT_PRIVATE_KEY": "0x..."
      }
    }
  }
}

Shortened. The full README is on GitHub.

Nothing above is checked by us. What we check is on the safety report.

Arbitova: common questions

Is Arbitova MCP server safe?
With care: it is graded C, so read the findings first (69/100). Read the Arbitova safety report
How do I install Arbitova?
It runs on your machine. Copy the Claude Code, Cursor, VS Code or Claude Desktop config from the install section.
Does Arbitova need an API key?
Yes. The registry entry asks for ARBITOVA_API_KEY.
Is Arbitova maintained?
The last commit was 190 days ago (2026-04-28). The latest release is v3.4.0.
How does Arbitova compare with other MCP servers?
safety score 69/100 (higher than 33% of 48,255 servers); tools 47 (the typical server has 7); tools that change things 26% (typical: 0%. Fewer write tools means less an agent can break). Full comparison.