deptrust CLI
Checks package versions for known vulnerabilities before install across npm, PyPI, Cargo, Go, Maven and more, with MCP and hooks
- 62 ★ on the repository
- Last commit
- MIT
Extended access
Why this level
- The installer changes Codex and Claude Code configuration and installs hooks, and checks call public registries and vulnerability databases.
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add cli/deptrustDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewExtended accessThe tool runs code or makes network calls and only reads external data.
- The installer changes Codex and Claude Code configuration and installs hooks, and checks call public registries and vulnerability databases.
This is third-party code. Review the repository files before installing.
What it does
`deptrust` queries the public OSV and GitHub Advisory Database sources and answers whether a package version is safe to install. It covers npm, PyPI, Cargo, Go modules, RubyGems, NuGet, Maven, Packagist, pub.dev, CocoaPods, Hex.pm, Hackage and GitHub Actions. The answer has a recommendation (`allow`, `review`, `block`), found vulnerabilities and risk signals: a version published in the last 72 hours, for example, goes to review. It runs as a CLI and as an MCP server, with no hosted service of its own. The installer wires MCP for Codex and Claude Code, installs a skill and `PreToolUse` hooks that check package install commands and GitHub Actions added to workflows before they run.
Who it is for. Developers whose agent picks dependency versions and installs packages, where checking happens before install, not after.
Good fit when
- The agent proposes dependency upgrades and you need to know which versions are safe
- A new package version is checked against vulnerability databases before it is added
- Hooks intercept `npm install` or `pip install` commands issued by the agent
- CI checks package versions and GitHub Actions used in workflows
Not a fit when
- You need analysis of your own source code, since the tool checks only known package advisories
- An `allow` result would be taken as proof a package is safe
- The environment has no access to public registries and OSV
Example request
Check `lodash` 4.17.20 and `vite` 7.0.0 with deptrust and say which to upgrade and which to block.Limitations
An `allow` result means only that no blocking known vulnerability was found, not that a package is safe. If no provider supports an ecosystem the answer is `unknown`. CocoaPods has no OSV data and Hackage has no GitHub Advisory Database. Hooks block a call on `review`, `block` and `unknown`, so false stops are possible. Unauthenticated GitHub requests hit rate limits; for CI the README recommends an app token in `DEPTRUST_GITHUB_TOKEN`.
How to disable. Run `npx @clidey/deptrust uninstall`, which removes the user-level binary, skill and MCP entries. Check hooks and tokens in the Codex and Claude Code configuration by hand.
FAQ
How do I install it?
Run `npx @clidey/deptrust install`. The installer asks which integrations to configure and shows the paths before changing anything. There is also `brew install clidey/tap/deptrust` and `go install github.com/clidey/deptrust/cmd/deptrust@latest`.
Which MCP tool is the main one?
`check_package` takes an ecosystem, package name and version (`latest` is allowed) and returns a compact answer with a recommendation, so the agent does not pull full advisory text into context.
How do I add MCP manually?
Run `deptrust mcp` as a stdio server. For Claude Code use `claude mcp add --transport stdio deptrust -- /absolute/path/to/deptrust mcp`.
What do the hooks do?
`PreToolUse` hooks check package install commands and GitHub Actions added to workflows, and block the call on `review`, `block` or `unknown`.
Is a GitHub token required?
No, though rate limits are lower without one. Locally you can set `DEPTRUST_GITHUB_AUTH=gh` to use the `gh` login, and no token is stored.
Discussion
No questions or reviews yet.
Related
CLI toolsSafeDep vet MCP: an agent checks npm and PyPI packages for malware, vulnerabilities and name squatting before install.
NPMScan plugin and MCP server: checks npm packages for malware and vulnerabilities, including install scripts and provenance
Security scanner for MCP servers that looks for hardcoded keys, tool poisoning and injection flaws in repository code.
The official Socket MCP server scores the security of npm, PyPI, cargo and other dependencies inside an agent chat.
An offline scanner vets repositories, skills, plugins and MCP servers for malicious code before they reach your agent.
The Snyk CLI for finding vulnerabilities in code, dependencies, containers and IaC, with a built-in MCP server and agent mode
README in short
The README covers supported ecosystems and providers, a coverage table, installs via `npx`, Homebrew, Go and Nix, agent setup, manual MCP wiring, the `check_package` tool and troubleshooting.