Test

Usage

snyk test [<OPTIONS>]

Description

The snyk test command checks projects for open-source vulnerabilities and license issues. The test command tries to auto-detect supported manifest files with dependencies and test those.

Note: There are specific snyk test commands for the Snyk Code, Container, and IaC scanning methods: code test, container test, and iac test.

Exit codes

Possible exit codes and their meaning:

0: success (scan completed), no vulnerabilities found 1: action_needed (scan completed), vulnerabilities found 2: failure, try to re-run the command. Use -d to output the debug logs. 3: failure, no supported projects detected

Configure the Snyk CLI

You can use environment variables to configure the Snyk CLI and set variables for connecting with the Snyk API. See Configure the Snyk CLI

Code execution warning

Before scanning your code, review the Code execution warning for Snyk CLI

Debug

Use the -d option to output the debug logs.

Options

See also subsequent sections for options for specific build environments, package managers, languages, and [<CONTEXT-SPECIFIC OPTIONS>] which you specify last.

--all-projects

Auto-detect all projects in the working directory, including Yarn workspaces.

For more information see the article Does the Snyk CLI support monorepos or multiple manifest files?

If you see the invalid string length error, refer to Invalid string length error when scanning projects

--fail-fast

Use with --all-projects to cause scans to be interrupted when errors occur and to report these errors back to the user.

The exit code is 2 and the scan ends. No vulnerability information is reported for projects that did not produce errors.

To perform the scan, resolve the error and scan again.

Note: If you do not use --fail-fast, Snyk scans all the projects but does not report any vulnerabilities for projects it could not scan due to misconfiguration or another error.

--detection-depth=<DEPTH>

Use with --all-projects or --yarn-workspaces to indicate how many subdirectories to search. DEPTH must be a number, 1 or greater; zero (0) is the current directory.

Default: no limit.

Example: --detection-depth=3 limits search to the specified directory (or the current directory if no <PATH> is specified) plus three levels of subdirectories; zero (0) is the current directory.

--exclude=<NAME>[,<NAME>]...>

Can be used with --all-projects and --yarn-workspaces to indicate directory names and file names to exclude. Must be comma-separated, and cannot include a path.

Example: $ snyk test --all-projects --exclude=dir1,file2

This will exclude any directories and files named dir1 and file2 when scanning for project manifest files such as: ./dir1, ./src/dir1, ./file2, ./src/file2 and so on.

Note: --exclude=dir1 will find both ./dir1, and ./src/dir1. However, --exclude=./src/dir1 will result in an error because it includes a path.

--prune-repeated-subdependencies, -p

Prune dependency trees, removing duplicate sub-dependencies.

Continues to find all vulnerabilities, but may not find all of the vulnerable paths.

Use this option if any big projects fail to be tested.

Default: false

--print-deps

Print the dependency tree before sending it for analysis.

--remote-repo-url=<URL>

Set or override the remote URL for the repository that you would like to monitor.

Groups all Projects found under a single Target.

--dev

Include development-only dependencies. Applicable only for some package managers, for example, devDependencies in npm or :development dependencies in Gemfile.

Note: This option can be used with Maven, npm, and Yarn projects.

Default: false, scan only production dependencies.

--org=<ORG_ID>

Specify the <ORG_ID> to run Snyk commands tied to a specific Snyk Organization. The <ORG_ID> influences some features availability and private test limits.

If you have multiple Organizations, you can set a default from the CLI using:

$ snyk config set org=<ORG_ID>

Set a default to ensure all newly tested projects are tested under your default Organization. If you need to override the default, use the --org=<ORG_ID> option.

Default: <ORG_ID> that is the current preferred Organization in your Account settings

Note: You can also use --org=<orgslugname>. The ORG_ID works in both the CLI and the API. The Organization slug name works in the CLI, but not in the API.

orgslugname must match the slug name as displayed in the URL of your org in the Snyk UI: https://app.snyk.io/org/[orgslugname]. The orgname does not work.

For more information see the article How to select the Organization to use in the CLI

--file=<FILE>

Specify a package file.

When you are testing locally or monitoring a project, you can specify the file that Snyk should inspect for package information. When the file is not specified, Snyk tries to detect the appropriate file for your project.

See also the section on Options for Python projects

--package-manager=<PACKAGE_MANAGER_NAME>

Specify the name of the package manager when the filename specified with the --file=<FILE> option is not standard. This allows Snyk to find the file.

Example: $ snyk test --file=req.txt --package-manager=pip

For more information see Options for Python projects

--unmanaged

For C++ only, scan all files for known open source dependencies.

For options you can use with --unmanaged see Options for scanning using --unmanaged

--ignore-policy

Ignore all set policies, the current policy in the .snyk file, Org level ignores, and the project policy on snyk.io.

--trust-policies

Apply and use ignore rules from the Snyk policies in your dependencies; otherwise ignore rules in the dependencies are only shown as a suggestion.

--show-vulnerable-paths=<none|some|all>

Display the dependency paths from the top level dependencies down to the vulnerable packages. Not supported with --json-file-output.

Default: some, a few example paths shown. false is an alias for none

Example: --show-vulnerable-paths=none

--project-name=<PROJECT_NAME>

Specify a custom Snyk project name.

--target-reference=<TARGET_REFERENCE>

Specify a reference that differentiates this project, for example, a branch name or version. Projects having the same reference can be grouped based on that reference. Supported for Snyk Open Source except for use with --unmanaged.

For more information see Group projects by branch or version for monitoring

You can use --target-reference=<TARGET_REFERENCE> when running tests to apply the same ignores and policies as for a monitored target.

For more information see Ignore issues

--policy-path=<PATH_TO_POLICY_FILE>

Manually pass a path to a .snyk policy file.

--json

Print results on the console as a JSON data structure.

Example: $ snyk test --json

If you see the invalid string length error, refer to Invalid string length error when scanning projects

--json-file-output=<OUTPUT_FILE_PATH>

Save test output as a JSON data structure directly to the specified file, regardless of whether or not you use the --json option.

Use to display the human-readable test output using stdout and at the same time save the JSON data structure output to a file.

For open source, Snyk creates a file whether or not issues are found. In contrast, for SAST, if no issues are found, Snyk does not create a json file.

Example: $ snyk test --json-file-output=vuln.json

If you see the invalid string length error, refer to Invalid string length error when scanning projects

--sarif

Return results in SARIF format.

--sarif-file-output=<OUTPUT_FILE_PATH>

Save test output in SARIF format directly to the <OUTPUT_FILE_PATH> file, regardless of whether or not you use the --sarif option.

This is especially useful if you want to display the human-readable test output using stdout and at the same time save the SARIF format output to a file.

--severity-threshold=<low|medium|high|critical>

Report only vulnerabilities at the specified level or higher.

--fail-on=<all|upgradable|patchable>

Fail only when there are vulnerabilities that can be fixed. Use one of the values as follows:

• all: Use to fail when there is at least one vulnerability that can be either upgraded or patched.

• upgradable: Use to fail when there is at least one vulnerability for which Snyk has a computed remediation available.

• patchable: Use to fail when there is at least one vulnerability that can be patched. Note that when you use patchable, the test will also fail if at least one vulnerability can be patched and other vulnerabilities found have a computed remediation available.

To fail on any Snyk-discoverable vulnerability (the default behavior), do not use the --fail-on option. If vulnerabilities do not have a Snyk-computed fix and this option is being used, tests pass.

Note: If you test code constrained by metadata that Snyk cannot respect with snyk test, Snyk will not propose a fix, in order to avoid breaking your code. You may be able to identify and apply a fix manually.

Options for Maven projects

Note: The --dev option can be used with Maven projects. See also the --dev option help

--maven-aggregate-project

Use --maven-aggregate-project instead of --all-projects when scanning Maven aggregate projects, that is, projects that use modules and inheritance.

Using --maven-aggregate-project instructs Snyk to perform a compilation step to ensure all modules within the project are resolvable by the Maven reactor. This ensures a comprehensive scan that includes dependencies of all sub-modules.

Be sure to run the scan in the same directory as the root pom.xml file.

Snyk reports the test results per individual pom.xml file within the aggregate project.

Note: You can use --all-projects when scanning Maven aggregate projects, but you cannot use --all-projects with --maven-aggregate-project.

--scan-unmanaged

To test individual JAR, WAR, and AAR files, use the following:

--scan-unmanaged --file=<JAR_FILE_NAME>

--scan-all-unmanaged

Auto-detect Maven, JAR, WAR, and AAR files recursively from the current folder.

--scan-all-unmanaged

Note: Custom-built JAR files, even with open-source dependencies, are not supported.

Options for Gradle projects

Note: If you see the invalid string length error, refer to Invalid string length error when scanning projects

--sub-project=<NAME>, --gradle-sub-project=<NAME>

For Gradle multi project configurations, test a specific sub-project.

--all-sub-projects

For multi project configurations, test all sub-projects. Both a build.gradle file and a settings.gradle file, or equivalent files, based on the package manager, must exist in the current directory.

--all-projects

See also the --all-projects option information in the Options section of this help.

Use for monorepos. This detects all supported manifests.

For Gradle monorepos Snyk looks only for root level build.gradle / build.gradle.kts files and applies the same logic as --all-sub-projects behind the scenes.

This option is designed to be run in the root of your monorepo.

For more details, see the following support article: Scanning Gradle projects in CLI with --exclude option does not exclude sub-projects

--configuration-matching=<CONFIGURATION_REGEX>

Resolve dependencies using the first configuration that matches the specified Java regular expression.

Example: ^releaseRuntimeClasspath$

--configuration-attributes=<ATTRIBUTE>[,<ATTRIBUTE>]...

Select certain values of configuration attributes to install and resolve dependencies.

Example: buildtype:release,usage:java-runtime

--init-script=<FILE>

Use for projects that contain a Gradle initialization script.

Options for NuGet projects

--assets-project-name

When you are monitoring a .NET project using NuGet PackageReference uses the project name in project.assets.json if found.

--file=<filename>.sln

Test all .NET projects included in the given .sln file. Projects referred to must have supported manifests.

Example: snyk test --file=myApp.sln

--file=packages.config

Test an individual .NET project.

--packages-folder

Specify a custom path to the packages folder.