exploiting-web3-smart-contracts

Fail

Audited by Gen Agent Trust Hub on Sep 16, 2026

Risk Level: CRITICALREMOTE_CODE_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [REMOTE_CODE_EXECUTION]: The skill facilitates the installation of the Foundry toolkit by piping a remote installation script from foundry.paradigm.xyz to the shell. Paradigm is a widely recognized and well-known Web3 research organization, and this is their official distribution channel for the tool.
  • [EXTERNAL_DOWNLOADS]: Environment configuration involves downloading a setup script for Node.js from Nodesource, a well-known service for official software binaries. The skill also references HackTricks, a well-known security research wiki, for blockchain vulnerability documentation.
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze untrusted smart contract source code and transaction data (ingestion points: Etherscan API and local Solidity files). The skill uses capabilities like subprocess calls to analysis tools (Slither, Mythril) to process this data. While explicit boundary markers are not used, this is standard practice for auditing skills where the code is treated as technical data for tool-based analysis rather than as a source of natural language instructions.
  • [COMMAND_EXECUTION]: The skill instructs the use of various command-line utilities for static analysis, fuzzing, and blockchain simulation, which are standard components of a smart contract auditor's toolkit.
Recommendations
  • HIGH: Downloads and executes remote code from: https://foundry.paradigm.xyz - DO NOT USE without thorough review
  • Contains 2 malicious URL(s) - DO NOT USE
Audit Metadata
Risk Level
CRITICAL
Analyzed
Sep 16, 2026, 11:06 AM
Security Audit — agent-trust-hub — exploiting-web3-smart-contracts