AuraOne Open / MCP security linting

Scan an MCP server before you install or ship it.

Statically lint MCP manifests, source, and docs for risky filesystem, shell, network, and auth patterns. Outputs Markdown, JSON, or SARIF for GitHub code scanning.

Input

  • An MCP server directory: manifest, package metadata, source, docs
  • Severity threshold for CI

Install and run

pip install mcp-risk-linter

The example inputs live in the repository. Get them first:

git clone --depth 1 https://github.com/auraoneai/mcp-risk-linter.git && cd mcp-risk-linter

Then run the example from the repository root:

mcp-risk-linter scan examples/safe_server --format json --fail-on high
- uses: auraoneai/[email protected]
  with:
    path: .
    format: sarif
    fail-on: none
Real outputJSON report for the bundled safe server: two read-only tools, no findings. mcp-risk-linter 0.1.6, captured 2026-11-20.
{
  "counts": {},
  "finding_count": 0,
  "findings": [],
  "root": "./examples/safe_server",
  "tools": [
    {
      "description": "Reads the local synthetic project summary metadata without writing files or calling external services.",
      "name": "read_project_summary",
      "source": "server.json"
    },
    {
      "description": "Lists static labels bundled with this synthetic fixture and has no side effects.",
      "name": "list_fixture_labels",
      "source": "server.json"
    }
  ]
}

Limitations

  • Static analysis only. It does not run the server, and it is not a full security audit.
  • Start with fail-on none and review false positives before enforcing in CI.

What you can do

Follow the work from input to a usable result.

Tool inventory

Lists every declared tool and where it was found.

Risk rules

Flags shell execution, broad filesystem access, network calls, and weak auth.

CI output

SARIF for GitHub code scanning, plus JSON and Markdown.

mcp-risk-linter workflow

See how the product moves from setup through a useful output.

  1. 01

    Install

    Install mcp-risk-linter 0.1.6 from the package registry.

    Package version

  2. 02

    Run on the bundled example

    Run the quick start on the synthetic fixture and compare it with the output above.

    Captured output

  3. 03

    Run on your data

    Point the same command at your own files, locally or in CI.

    Local report

Release specifications

What is included in this release

Review the current source, release, runtime, and support information.

Package
mcp-risk-linter 0.1.6Discovery snapshot registry readback
Source tag
v0.1.6 (7fc193dacee8)Public GitHub release tag
Sample
examples/safe_server, captured 2026-11-20public/open/samples.json
Data handling
Scanning runs inside the local or CI repository boundary and requires no model call.Project README

Availability

Current release and install details.

Review the current version, source, packages, downloads, supported platforms, and install options.

Verified

mcp-risk-linter availability

PyPI 0.1.6 is available with a linked source tag v0.1.6.

When the work grows

See related agent evaluation and reliability workflows.

Talk to AuraOne