A map an agent can read in one pass.

# ChainRay

> 57-chain on-chain intelligence and agent infrastructure for discovery, trust, safety, and execution.
> 57 chains (44 EVM + 13 non-EVM). 93 logical paid capabilities / 94 HTTP route templates + 2 free HTTP surfaces. 79 MCP tools.
> Chain Proof v3 (dataHash + analysisHash). A2A + KYA + Safety Analysis + Solver Feed + Sentiment + CaaS.
> Pay-per-request with USDC via x402. No API keys needed.

## Start here

- [Agent documentation](https://chainray.online/docs): Human-readable API reference and workflow guidance.
- [OpenAPI contract](https://chainray.online/openapi.json): Machine-readable HTTP route and payment schemas.
- [MCP server card](https://chainray.online/.well-known/mcp/server-card.json): Streamable HTTP connection metadata.
- [MCP capability map](https://chainray.online/.well-known/mcp-capabilities.json): Core and specialist MCP tools, aliases, composites, and HTTP handoffs.
- [Authentication and payment model](https://chainray.online/auth.md): Public access, x402 settlement, and client-side signing rules.

## MCP and HTTP coverage

- MCP exposes 79 top-level tools: 38 core tools plus 41 specialist tools, including all 78 legacy names.
- The paid HTTP contract exposes 93 logical capabilities across 94 route templates.
- All 78 legacy MCP tools remain directly callable alongside the new-pair discovery tool. Composite views do not replace their focused components.
- Some HTTP-specific capabilities, including live SSE streams, require a direct HTTP connection; util.stream_info provides connection guidance.
- Use the capability map and OpenAPI contract before concluding that a capability is unavailable from the service.

## v1.0 New Endpoints — Agent Economy

### A2A Agent Card ($0.002-$0.05) — Agent Discovery & Registration:
- POST /agent-card/register ($0.05): Register an AI agent in the A2A registry
- /agent-card/discover ($0.007): Discover agents by capability, chain, or keyword
- /agent-card/{agentId} ($0.002): Look up a specific agent's card

### Pre-Trade Safety Analysis ($0.05) — Risk Assessment:
- /mev-shield ($0.05): 5-factor sandwich attack risk analysis with slippage/mitigation advice (heuristic model, no mempool access)

### ⚡ Intent Solver Feed ($0.007) — Execution Intelligence:
- /solver-feed ($0.007): Cross-chain profitability analysis for ERC-7683 intent solvers

### 📊 Token Sentiment Oracle ($0.012) — Market Sentiment:
- /sentiment/{token} ($0.012): 5-signal on-chain sentiment analysis (zero LLM)

### 🧠 CaaS v2 Workflow Brief ($0.05) — Multi-Agent Orchestration:
- POST /context-brief/workflow ($0.05): Role-based context briefs for collaborative agent workflows on up to three EVM chains; unsupported non-EVM selections are rejected before payment

### 🔮 Data Oracle ($0.002-$0.007) — Verified Feeds:
- /oracle/gas-avg ($0.007): Multi-chain gas average
- /oracle/price-feed ($0.002): Verified price oracle
- /oracle/chain-status ($0.007): Multi-chain health status

### 🛡️ KYA Agent Reputation ($0.025) — Trust Layer:
- /agent-reputation/{address} ($0.025): 10-factor trust scoring (ERC-8004, ?format=erc8004)

### 📊 CaaS v1 Context Brief ($0.05) — Context Layer:
- /context-brief ($0.05): Multi-signal context aggregation

## Current Pricing Tiers

- $0.002: 11 paid endpoints
- $0.005: 2 paid endpoints
- $0.007: 15 paid endpoints
- $0.01: 1 paid endpoints
- $0.012: 18 paid endpoints
- $0.015: 1 paid endpoints
- $0.02: 2 paid endpoints
- $0.025: 21 paid endpoints
- $0.03: 1 paid endpoints
- $0.035: 5 paid endpoints
- $0.04: 3 paid endpoints
- $0.05: 12 paid endpoints
- $0.08: 1 paid endpoints

(See /openapi.json for full endpoint list)

## Features
- Chain Proof v3: dataHash + analysisHash (verifiable AI outputs)
- Address Intelligence: Auto-learning whale/smart money database
- A2A Agent Card Registry: Google A2A protocol compatible
- ERC-8004 Reputation Registry: Structured trust data
- Pre-trade safety, cross-chain decision context, wallet/protocol intelligence, machine-verifiable evidence, and agent identity/trust
- Privacy-first: No-log, no-track, no-KYC
- MCP: /mcp — 79 tools for AI agents (legacy SSE: /mcp/sse)

## Discovery
- [x402 resource discovery](https://chainray.online/.well-known/x402.json): Paid HTTP resources, prices, chain support, and payment metadata.
- [A2A Agent Card](https://chainray.online/.well-known/agent-card.json): Agent identity and workflow capabilities.
- [Agent Skills index](https://chainray.online/.well-known/agent-skills/index.json): Signed-by-content workflow guides.
- [API Catalog](https://chainray.online/.well-known/api-catalog): RFC 9264 linkset for service contracts and status.
- [AI Catalog](https://chainray.online/.well-known/ai-catalog.json): Agent-facing capability summaries and representative queries.
- [Privacy policy](https://chainray.online/.well-known/privacy.json): No-log, no-track, no-KYC policy.
- [MCP endpoint](https://chainray.online/mcp): MCP tool discovery and Streamable HTTP transport.
- [Complete endpoint reference](https://chainray.online/llms-full.txt): Full route and pricing reference.

## Agent workflows

- [Pre-trade safety](https://chainray.online/agent-skills/pre-trade-safety.md): Token, contract, approval, liquidity, and MEV signals before action.
- [Cross-chain decision context](https://chainray.online/agent-skills/cross-chain-decision-context.md): Gas, chain health, bridges, liquidity, and solver context.
- [Wallet and protocol intelligence](https://chainray.online/agent-skills/wallet-protocol-intelligence.md): Wallet activity, behavior, protocol exposure, and trust context.
- [Machine-verifiable evidence](https://chainray.online/agent-skills/machine-verifiable-evidence.md): Proof hashes, freshness, block context, and verification.
- [Agent identity and trust](https://chainray.online/agent-skills/agent-identity-and-trust.md): Agent cards, payment identity, reputation, and evidence boundaries.

## Contact
- Email: chainraydev@proton.me