web3-solidity-audit-mcp

Fail

Audited by Gen Agent Trust Hub on Aug 27, 2026

Risk Level: HIGHEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [REMOTE_CODE_EXECUTION]: The installation instructions utilize shell piping (curl | bash) to execute remote setup scripts for the Foundry toolchain and the Aderyn static analyzer. These scripts are sourced from well-known technology organizations.
  • [EXTERNAL_DOWNLOADS]: The skill requires the installation of various external dependencies, including slither-analyzer, solc-select, and halmos via Python's package manager. It also installs the solidity-audit-mcp package globally from NPM, which is hosted on a third-party GitHub repository (mariano-aguero/solidity-audit-mcp). These packages are installed without version pinning, which can introduce supply chain risks.
  • [INDIRECT_PROMPT_INJECTION]:
  • Ingestion points: Untrusted smart contract files are ingested through parameters such as contractPath and projectRoot across multiple tools like analyze_contract and audit_project.
  • Boundary markers: No delimiters or instructions are provided to separate the ingested code from the agent's logic, creating a surface where malicious code comments could attempt to influence the agent's audit results or behavior.
  • Capability inventory: The skill can execute subprocesses for analysis and testing, read project files, and generate exploit templates or reports based on the contract content.
  • Sanitization: There is no evidence of sanitization or validation of the input Solidity code prior to its analysis by the integrated tools or the LLM.
Recommendations
  • HIGH: Downloads and executes remote code from: https://foundry.paradigm.xyz, https://raw.githubusercontent.com/Cyfrin/aderyn/dev/cyfrinup/install - DO NOT USE without thorough review
Audit Metadata
Risk Level
HIGH
Analyzed
Aug 27, 2026, 05:38 AM
Security Audit — agent-trust-hub — web3-solidity-audit-mcp