pysnyk-mcp 1.2.0

Model Context Protocol (MCP) server exposing Snyk via pysnyk

snyk-mcp

A Model Context Protocol (MCP) server that exposes Snyk to LLM clients (Claude Desktop, Claude Code, Cursor, VS Code, etc.) via the official pysnyk SDK.

Ask an LLM things like "which of our npm projects have critical vulnerabilities?", "show me all CVEs in our nginx container image", or "test the latest version of lodash for known CVEs" — and the LLM answers by calling Snyk directly through this server.

PyPI: pip install pysnyk-mcp  ·  Image: ghcr.io/polarpoint-io/snyk-mcp:latest  ·  Repo: https://github.com/polarpoint-io/snyk-mcp

Table of contents

• Quick start
• Configuration
• Client integrations
• Tool reference
• Example prompts
• Running from source
• Development
• Publishing
• Security notes

Quick start

1. pip (no Docker)

pip install pysnyk-mcp
export SNYK_TOKEN=your_snyk_api_token

pysnyk-mcp                                    # stdio
TRANSPORT=http HTTP_PORT=8000 pysnyk-mcp      # HTTP/SSE

2. Docker (stdio — launched by your MCP client)

docker pull ghcr.io/polarpoint-io/snyk-mcp:latest

docker run --rm -i \
  -e SNYK_TOKEN=your_snyk_api_token \
  ghcr.io/polarpoint-io/snyk-mcp:latest

3. Docker (HTTP/SSE — standalone service)

docker run --rm \
  -p 8000:8000 \
  -e SNYK_TOKEN=your_snyk_api_token \
  -e TRANSPORT=http \
  ghcr.io/polarpoint-io/snyk-mcp:latest
# SSE endpoint: http://localhost:8000/sse

Get your token from: Snyk → Account Settings → Auth Token at https://app.snyk.io/account.

Configuration

Variable	Required	Default	Description
SNYK_TOKEN	yes	—	Snyk API token
SNYK_API_URL	no	https://api.snyk.io/v1	Override API base URL
TRANSPORT	no	stdio	stdio or http
HTTP_HOST	no	0.0.0.0	Bind host (HTTP mode)
HTTP_PORT	no	8000	Bind port (HTTP mode)
LOG_LEVEL	no	INFO	Python log level

See .env.example for a copy-pasteable template.

Client integrations

Each client supports two transport options — pip (recommended, no Docker required) or Docker.

# Install once
pip install pysnyk-mcp

---

Claude Desktop

Edit ~/Library/Application Support/Claude/claude_desktop_config.json (macOS) or %APPDATA%\Claude\claude_desktop_config.json (Windows).

pip (recommended)

{
  "mcpServers": {
    "snyk": {
      "command": "pysnyk-mcp",
      "env": {
        "SNYK_TOKEN": "your_snyk_api_token"
      }
    }
  }
}

Docker

{
  "mcpServers": {
    "snyk": {
      "command": "docker",
      "args": [
        "run", "--rm", "-i",
        "-e", "SNYK_TOKEN",
        "ghcr.io/polarpoint-io/snyk-mcp:latest"
      ],
      "env": {
        "SNYK_TOKEN": "your_snyk_api_token"
      }
    }
  }
}

Restart Claude Desktop after editing. You should see a hammer icon in the chat confirming the snyk server is connected with all 23 tools available.

---

Claude Code

pip (recommended)

pip install pysnyk-mcp

claude mcp add snyk -- pysnyk-mcp
# then export your token before running claude:
export SNYK_TOKEN=your_snyk_api_token

Docker

claude mcp add snyk -- docker run --rm -i \
  -e SNYK_TOKEN \
  ghcr.io/polarpoint-io/snyk-mcp:latest

export SNYK_TOKEN=your_snyk_api_token

---

Cursor

Edit ~/.cursor/mcp.json:

pip (recommended)

{
  "mcpServers": {
    "snyk": {
      "command": "pysnyk-mcp",
      "env": {
        "SNYK_TOKEN": "your_snyk_api_token"
      }
    }
  }
}

Docker

{
  "mcpServers": {
    "snyk": {
      "command": "docker",
      "args": ["run", "--rm", "-i", "-e", "SNYK_TOKEN",
               "ghcr.io/polarpoint-io/snyk-mcp:latest"],
      "env": { "SNYK_TOKEN": "your_snyk_api_token" }
    }
  }
}

---

VS Code (Continue)

Edit ~/.continue/config.json:

pip (recommended)

{
  "mcpServers": [{
    "name": "snyk",
    "command": "pysnyk-mcp",
    "env": { "SNYK_TOKEN": "your_snyk_api_token" }
  }]
}

Docker

{
  "mcpServers": [{
    "name": "snyk",
    "command": "docker",
    "args": ["run", "--rm", "-i", "-e", "SNYK_TOKEN",
             "ghcr.io/polarpoint-io/snyk-mcp:latest"],
    "env": { "SNYK_TOKEN": "your_snyk_api_token" }
  }]
}

---

HTTP / SSE (remote or shared server)

If you prefer to run the server as a persistent HTTP service rather than a subprocess:

# pip
pip install pysnyk-mcp
SNYK_TOKEN=your_token TRANSPORT=http HTTP_PORT=8000 pysnyk-mcp

# Docker
docker run --rm -p 8000:8000 \
  -e SNYK_TOKEN=your_snyk_api_token \
  -e TRANSPORT=http \
  ghcr.io/polarpoint-io/snyk-mcp:latest

Then point your MCP client at http://localhost:8000/sse.

Tool reference

The server exposes 23 tools grouped by capability. All return JSON. Errors from the Snyk API are caught and returned as {"error": "...", "message": "..."} so the LLM can handle them gracefully.

Discovery (start here)

Tool	Description
snyk_list_org_ids	Compact org name → ID listing — use to find org_id values
snyk_list_project_ids	Compact project name → ID listing across orgs — use to find project_id values

Organizations

Tool	Description
snyk_list_organizations	List all orgs the token has access to (full details)
snyk_get_organization	Get details for a single org by ID

Projects

Tool	Description
snyk_list_projects	List projects with full details (filter by org, name, or origin)
snyk_get_project	Full details for a single project
snyk_get_project_dependency_graph	Full dependency tree for a project

Issues & Vulnerabilities

Tool	Description
snyk_list_project_vulnerabilities	Open vulnerabilities with optional severity filter
snyk_list_aggregated_issues	Deduplicated issues with fix info (recommended)

CVEs

Tool	Description
snyk_list_project_cves	Flat CVE list (CVE ID, CVSS score, package, upgrade path) for any project, sorted by severity
snyk_list_container_projects	Discover container image projects in an org (Docker Hub, ECR, ACR, GCR, Quay, Harbor, …)
snyk_list_container_image_cves	CVEs for a specific container image project

Dependencies & Licenses

Tool	Description
snyk_list_dependencies	Packages in use (org-wide or per project)
snyk_list_org_licenses	Licenses across an org — compliance audit
snyk_list_project_licenses	Licenses for a specific project

Members

Tool	Description
snyk_list_members	Members of a Snyk organization

Integrations

Tool	Description
snyk_list_integrations	Active integrations (GitHub, Docker Hub, npm, etc.)

Package Testing

Tool	Description
snyk_test_npm_package	Test an npm package for known vulnerabilities
snyk_test_python_package	Test a PyPI package for known vulnerabilities
snyk_test_maven_package	Test a Maven artifact for known vulnerabilities
snyk_test_rubygem_package	Test a Ruby gem for known vulnerabilities

Tags

Tool	Description
snyk_list_project_tags	Tags applied to a project (key/value pairs)

User / Health

Tool	Description
snyk_get_user_info	Authenticated user details — useful as a connectivity check

Example prompts

Security triage: "List all projects with critical vulnerabilities in our main org."

Package audit: "Is the version of Flask we're using (2.3.0) affected by any known CVEs?"

CVE report: "Give me all critical and high CVEs in the payment-service project with their CVSS scores and upgrade paths."

Container audit: "List all our container image projects and show me any critical CVEs in the nginx image."

Log4Shell check: "Do any of our container images have CVE-2021-44228?"

License compliance: "Show me all high-severity license issues across our organization."

Maven check: "Test log4j-core 2.14.1 — is it affected by Log4Shell?"

Inventory: "How many projects do we have per origin (github, cli, docker)?"

Member audit: "Who are the admins in our Snyk org?"

Running from source

git clone [e-mail]:polarpoint-io/snyk-mcp.git
cd snyk-mcp
python -m venv .venv && source .venv/bin/activate
pip install -e .[dev]

export SNYK_TOKEN=...

pysnyk-mcp                                    # stdio
TRANSPORT=http HTTP_PORT=8000 pysnyk-mcp      # HTTP/SSE

Or via the Makefile:

make dev       # install with dev extras
make test      # pytest
make lint      # ruff check
make run       # stdio
make run-http  # HTTP/SSE on :8000
make docker    # build the image locally

Development

make dev
make test
make lint

Adding a new tool:

1. Add a @mcp.tool() function in src/pysnyk_mcp/server.py, wrapping the pysnyk call in a _run() closure.
2. Pass _run to _safe_call() to ensure exceptions return cleanly as {"error": ..., "message": ...}.
3. Add a mocked-client test in tests/test_tools.py.

See AGENTS.md for deeper contribution guidance.

Publishing

Releases are fully automated via semantic-release — no manual tagging needed.

How it works

Every push to main runs the CI pipeline:

1. Test — lint (ruff) + pytest across Python 3.10 / 3.11 / 3.12
2. Docker — builds and pushes the image to GHCR (multi-arch amd64/arm64)
3. Release — semantic-release analyses conventional commits, bumps the version, updates CHANGELOG.md, publishes to PyPI, and creates a GitHub Release

A release only happens when commits contain a feat:, fix:, or breaking-change — chore:, docs:, ci: commits don't trigger one.

Docker image tags

Tag	When pushed
latest	Every merge to main
sha-<short>	Every merge to main
1.2.3 / 1.2	On a semantic-release version bump

Required GitHub secrets

Secret	Description
POL_GH_TOKEN	Personal access token with repo + write:packages scope
PYPI_TOKEN	PyPI API token for the pysnyk-mcp project

Add both at: GitHub repo → Settings → Secrets and variables → Actions.

Commit conventions

feat: add new tool          → minor version bump (0.x.0)
fix: correct response shape → patch bump        (0.0.x)
feat!: breaking change      → major bump        (x.0.0)
chore/docs/ci/test          → no release

Security notes

• The container runs as a non-root user (uid 10001).
• SNYK_TOKEN is read from env vars only — never written to disk.
• Destructive operations (project.delete(), project.move()) are deliberately not exposed.
• If you add write tools, guard them behind a separate env flag (e.g. SNYK_ENABLE_WRITE=1).

License

MIT — see LICENSE.