hound-mcp

Hound MCP Server

Hound MCP brings dependency checks into AI coding workflows, helping agents and developers evaluate packages before installation and review existing projects. Its tools cover known vulnerabilities, license policies, dependency trees, and suspected typosquatting.

Server facts

hound-mcp

Project

License
Not specified

Tech

Runtime / language
Not specified
Transports
stdio
Remote available
No
Authentication
Not specified

Activity

GitHub stars
17
Last commit
Data as of

Overview

Package selection and lockfile review are the focus of Hound MCP’s dependency security workflow. Before adding a dependency, agents can use hound_preinstall to assess vulnerability reports, suspicious naming, abandonment, and licensing concerns. Package comparisons weigh security findings alongside maintenance signals and licenses, while hound_score combines several signals into a graded assessment. For closer investigation, package profiles expose OpenSSF Scorecard information and repository statistics; dependency trees reveal transitive relationships; and advisory lookup accepts GHSA, CVE, or OSV identifiers. Typosquatting checks generate plausible name variants and look for registry matches.

For existing projects, hound_audit checks dependencies parsed from supported lockfiles against OSV vulnerability data. License checks flag dependencies that conflict with a selected policy, and hound_upgrade searches published versions for the nearest upgrade addressing known vulnerabilities. Built-in prompts organize project audits, dependency evaluations, and checks before release. Hound uses public deps.dev and OSV APIs without requiring API keys. It connects through stdio to MCP clients, with setup examples for Claude Code, Cursor, Windsurf, and VS Code.

hound-mcp setup notes

  • hound-mcp records stdio only; confirm a network transport upstream before planning remote access.
  • hound-mcp has no recorded license; confirm reuse terms in its repository.

Related MCP servers