Etherscan launches Transaction Debugger skill for AI agents

Etherscan has released a Transaction Debugger skill that turns complicated Ethereum transactions into reports, either simple or detailed, using its own data. The block explorer designed the tool for agent tools such as Claude Code and Codex, giving them a structured way to analyze EVM transactions with live information from Etherscan, rather than relying on whatever the model thinks it remembers about a contract.

What the Transaction Debugger actually does

The debugger decodes the calls inside a transaction and lays out the path execution actually took. It also reconciles changes to assets and permissions, tracking who ended up with what, and which contracts gained the right to spend which tokens. When a transaction fails, the skill offers failure analysis.

According to Etherscan’s materials, the output includes a summary of the transaction, the execution flow, and security observations. Report modes span from Simple to Security. Each report includes direct Etherscan links, and Etherscan positions every output as a read-only reconstruction drawn from its API that users should verify through the explorer.

The skill can examine a single transaction or compare two side by side. It works on any EVM chain that Etherscan supports.

Part of a bigger Build with AI push

The Transaction Debugger belongs to Etherscan’s Build with AI suite, which launched in August 2026. Public promotion of the debugger began on September 4, 2026, and additional guidance followed on October 16, 2026.

Etherscan’s MCP server provides structured, read-only access to on-chain data across more than 60 EVM-compatible chains. The company has stressed pairing the debugger with either its CLI or MCP server so agents can pull data smoothly.

Two companion skills fill out the toolkit. Etherscan Flow tracks the movement of funds, while Contract Review is aimed at validating contracts. The debugger intentionally avoids broad wallet tracing and multi-hop fund flows, handing those jobs to Etherscan Flow.

Installation is a single command for developers already working in agent environments:

npx skills add https://github.com/etherscan/skills –skill etherscan-transaction-debugger -g -a ‘*’ -y

Why grounding the AI matters

Etherscan emphasizes that the debugger’s outputs come straight from its API as a guard against hallucinated results. The agent produces the narrative, but the explorer remains the source of truth that users are expected to check.