The AI-native block explorer
Explorers today show data; OpenScan.AI explains it. Our intelligence layer turns raw on-chain noise into clear, actionable answers — live today on XDC Network.
From calldata to clarity
A real XDC transaction, two ways. On the left, what explorers show today; on the right, what the Phase 3 AI decoder will tell you.
method: transferFrom(address,address,uint256)
raw_input: 0x23b872dd
0000000000000000000000004c297e439525063bfd4a1dff018c84e317d7e25e
000000000000000000000000af9b78d4da806e4f0f7dc37d3f40b0e62683d15e
00000000000000000000000000000000000000000000000000000000223faaaaTransfer of 572.201642 USDC from 0x4c29…7e25e to 0xaf9b…d15e via the USDC token contract. The recipient initiated the transfer, pulling tokens they were previously approved to spend.
Illustrative preview of the AI transaction decoder — ships in Phase 3 (Q3 2026).
View this transaction on xdcscan.io →Try it: ask about a transaction
Paste any XDC transaction hash. You’ll get the real on-chain data plus a preview of the plain-English summary our Phase 3 AI decoder will generate for every transaction.
On-chain data (example)
- Hash
- 0xffdc188ae9be6610…
- From
- 0xaf9b…d15e
- To
- 0xfa29…8eb1
- Value
- 0 XDC
- Method
- transferFrom
- Block
- 106,754,420
- Timestamp
- 2026-09-02T10:47:36Z
AI summary (preview)
Transfer of 572.201642 USDC from 0x4c29…7e25e to 0xaf9b…d15e via the USDC token contract. The recipient initiated the transfer, pulling tokens they were previously approved to spend.
Template-generated preview of the Phase 3 AI transaction decoder — not a live LLM call. Full AI decoding ships in Q3 2026.
What “AI-powered” actually means
Every capability below is mapped to a concrete roadmap phase — no vaporware claims. Badges show exactly when each ships. Building an autonomous agent? Start on the agents page.
AI Transaction Decoding
Coming Q3 2026Raw calldata and event logs become human-readable summaries — see what a transaction actually did, not just its hash. Ships as the AI transaction decoder in Phase 3.
Phase 3 roadmap →Behavior Analysis
Coming Q3 2026Understand how wallets and contracts behave over time, with patterns and interactions surfaced automatically instead of buried in raw logs.
Predictive Insights & Anomaly Detection
Coming Q3 2026Surface anomalies, trends, and network-level signals before they matter to your users or your ecosystem — part of the Phase 3 intelligence layer alongside automated security scanning.
Phase 3 roadmap →MCP Server for AI Agents
Open sourceThe OpenScan.AI MCP server wraps the explorer APIs and exposes balances, tokens, NFTs, and contract metadata to AI agents and IDEs like Claude and Cursor. Self-host it and point it at any OpenScan.AI instance.
MCP server docs →Natural-Language Search
VisionThe long-term vision: ask the explorer questions in plain English — “show me large USDC movements this week” — and get answers, not raw JSON. Built on the Phase 3 intelligence layer.
Connect your AI agent
The OpenScan.AI MCP server wraps the explorer API so AI agents, IDEs, and automation tools (Claude, Cursor, and other MCP hosts) can query XDC data contextually. Self-host it from GitHub and point it at any OpenScan.AI instance.
{
"mcpServers": {
"openscan": {
"command": "<command-to-run-your-mcp-server>",
"env": {
"OPENSCAN_API_URL": "https://xdcscan.io/api/v2"
}
}
}
}What your agent can query
The server supports configurable explorer URLs, custom host instructions, and progress notifications for long-running queries.
Building the intelligence layer in the open
AI transaction decoding, security scanning, and wallet visualization all land in Phase 3. Read the plan, or see it in action on XDC today.