codeql

Subscribe to all “codeql” posts via RSS or follow GitHub Changelog on Twitter to stay updated on everything we ship.

→ ~ cd github-changelog
→ ~/github-changelog|main git log main
showing all changes successfully

Code scanning shows more accurate and relevant alerts on pull requests

Code scanning is now using a new way of analysing and displaying alerts on pull requests. The change ensures code scanning only shows accurate and relevant alerts for the pull request.

Previously, code scanning presented all alerts unique to the pull request branch, even if they were unrelated to the code changes the pull request introduced. Now, the tool reports only alerts inside the lines of code that the pull request has changed, which makes it easier to fix these contextualised alerts in a timely manner.

code scanning on the slide-out enablement panel on the security coverage page

The complete list of code scanning alerts on the pull request branch can be seen on the Security tab of the repository.

code scanning on the slide-out enablement panel on the security coverage page

In addition, code scanning will no longer show fixed alerts on pull requests. Instead, you can check whether an alert has been fixed by your pull request on the Security tab of the repository by using search filters: pr:111 tool:CodeQL. If you fix an alert in the initial commit in the pull request, it will not be present on the PR branch.

This has shipped to GitHub.com and will be available in GitHub Enterprise Server 3.10.

Learn more about viewing an alert on your pull request.

Learn more about GitHub Advanced Security.

See more

Use multi-repository variant analysis (beta) to run CodeQL queries at scale

Today we have released multi-repository variant analysis for CodeQL in public beta to help the OSS security community power up their research with CodeQL.

CodeQL is the static code analysis engine that powers GitHub code scanning. Out of the box, CodeQL is able to find many different types of security vulnerability and flag them up in pull requests.

But one of CodeQL’s superpowers is its versatility and customizability: you can use it to find virtually any pattern in source code. As such, it’s a great tool for finding new types of vulnerabilities – once you’ve identified an interesting pattern, model it as a CodeQL query, and then run it against your repository to find all occurrences of that pattern! But most vulnerabilities are relevant to many codebases. Wouldn’t it be amazing if you could easily run your query against many repos at the same time? Well, now you can with multi-repository variant analysis — which we’ve just shipped in public beta!

Screenshot 2023-02-22 at 16 39 39

This new feature will allow security researchers to run CodeQL analyses against large numbers of repos, straight from the CodeQL extension for VS Code, making it possible to identify new types of security vulnerabilities in the most popular open-source codebases.

Checkout the CodeQL for VS Code documentation to get learn how to get started with multi-repository variant analysis. We'd also love to hear your feedback on this GitHub community discussion.

See more

Code scanning adds support for Kotlin 1.8.0 analysis (beta)

Back in November 2022 we announced the public beta for Kotlin analysis. We continue to invest in Kotlin and we now support Kotlin 1.8.0 analysis in beta.

If you have any feedback or questions, please use this discussion thread or open an issue in the open source CodeQL repository if you encounter any problems.

Kotlin beta support is available by default in GitHub.com code scanning, the CodeQL CLI, and the CodeQL extension for VS Code. GitHub Enterprise Server (GHES) version 3.9 will include this beta release.

See more

CodeQL code scanning is now 16% faster

CodeQL is the engine that powers GitHub code scanning, used by more than 100,000 repositories to catch security vulnerabilities before they cause issues in deployments.

CodeQL is fully integrated into the Pull Request workflow, so it has to be as fast as possible to keep developers unblocked.

We're constantly working on performance improvements, from incremental optimizations to fundamental research, all with the goal of speeding up the nearly 150,000 checks we run every single day, without compromising our best-in-class precision and low false-positive rate.

With the recent release of CodeQL version 2.12, we looked back at the performance gains compared to version 2.11 (September 2022) to see how far we've come. We compared the analysis time for the same 55,000 repositories on GitHub.com and found an average improvement of 15.7% across all supported languages:

codeql performance 2 11 2 12 improvement

Users on GitHub.com automatically run the latest CodeQL version. Customers on GitHub Enterprise Server can update by following the sync processes explained here.

See more

Migration of CodeQL CLI Docs to GitHub Docs site

Following feedback from code scanning users, we've moved documentation about the CodeQL CLI from codeql.github.com to docs.github.com, the main GitHub Docs site.

You can now find the articles under the “Using the CodeQL CLI” and “CodeQL CLI reference” categories, which correspond to the categories on the original site. We’ve updated each of the original articles on codeql.github.com with links to the new location of the article and to each subsection, so that if you go to the old location you can easily find the information you need.

The source files now exist in Markdown format in the public, open-source docs repository. If you would like to contribute, you can consult and follow the steps listed in the GitHub Docs contributing guide.

See more

Code scanning can be set up not to fail a pull request check

Code scanning can now be set up to never cause a pull request check failure.

By default, any code scanning alerts with a security-severity of critical or high will cause a pull request check failure.
You can specify which security-severity level for code scanning results should cause the code scanning check to fail, including None, by going to the Code security and Analysis tab in the repository settings.

Screenshot code-scanning-settings

This has shipped to GitHub.com and will be available in GitHub Enterprise Server 3.9. Learn more about severity levels for security alerts and Code scanning results check failures on pull requests.

See more

Code scanning: CodeQL Action v1 is now deprecated

On March 30, 2022, we released CodeQL Action v2, which runs on the Node.js 16 runtime. In April 2022, we announced that CodeQL Action v1 would be deprecated at the same time as GitHub Enterprise Server (GHES) 3.3.
This deprecation period has elapsed and starting January 18, 2023, CodeQL Action v1 is now discontinued.
It will no longer be updated or supported, and while we will not be deleting it except in the case of a security vulnerability, workflows using it may eventually break.
New CodeQL analysis capabilities will only be available to users of v2.

For more information about this deprecation, please see the original deprecation announcement from April 2022.

How does this affect me?

If you use code scanning with CodeQL on any of the following platforms, you should update your workflow file(s) to use CodeQL Action v2 as soon as possible:

  • GitHub.com (including open source repositories, users of GitHub Teams and GitHub Enterprise Cloud)
  • GHES 3.4.13 and later

Users of GHES 3.4.12 or earlier: please read this section in the original deprecation announcement.

What do I need to change in my workflow?

To upgrade to the CodeQL Action v2, open your CodeQL workflow file(s) in the .github/workflows directory of your repository and look for references to:

  • github/codeql-action/init@v1
  • github/codeql-action/autobuild@v1
  • github/codeql-action/analyze@v1
  • github/codeql-action/upload-sarif@v1

These entries need to be replaced with their v2 equivalents:

  • github/codeql-action/init@v2
  • github/codeql-action/autobuild@v2
  • github/codeql-action/analyze@v2
  • github/codeql-action/upload-sarif@v2

If you use a pinned version of the CodeQL Action in your workflows, for example github/codeql-action/init@32be38e, check the latest Actions workflow run summary on your repository.
If you see a warning stating that you are running CodeQL Action v1, then please update your workflow to reference v2 or alternatively the latest github/codeql-action commit tagged v2.

Can I use Dependabot to help me with this upgrade?

All users on GitHub.com, and GHES customers using GitHub Advanced Security with a local copy of github/codeql-action, can use Dependabot to automatically upgrade their Actions dependencies.
For more details on how to set this up, please see this page.

GHES customers should also make sure:

See more

Updated timeline for the deprecation of CodeQL Action v1

The deprecation date for the CodeQL Action v1 is shifting. Initially, this was December 2022, and now it is January 2023. This change follows the updated timeline on the deprecation of GitHub Enterprise Server (GHES) 3.3.

In January 2023, the CodeQL Action v1 will be officially deprecated (alongside GHES 3.3). GitHub Action workflows that refer to v1 of the CodeQL Action will continue to work, but no new analysis capabilities will be released to v1. New CodeQL analysis capabilities will only be available to users of v2. For more information about this deprecation and detailed upgrade instructions, please see the original deprecation announcement from April 2022.

All users of GitHub code scanning (which by default uses the CodeQL analysis engine) on GitHub Actions on the following platforms should update their workflow files:

Environments in which CodeQL runs in CI/CD systems other than GitHub Actions are not affected by this deprecation.

See more

CodeQL code scanning launches Kotlin analysis support (beta)

Starting today, GitHub code scanning includes beta support for analyzing code written in Kotlin, powered by the CodeQL engine.

Kotlin is a key programming language used in the creation of Android mobile applications, and is an increasingly popular choice for new projects, augmenting or even replacing Java. To help organisations and open source developers find potential vulnerabilities in their code, we’ve added Kotlin support (beta) to the CodeQL engine that powers GitHub code scanning. CodeQL now natively supports Kotlin, as well as mixed Java and Kotlin projects. Set up code scanning on your repositories today to receive actionable security alerts right on your pull-requests. To enable Kotlin analysis on a repository, configure the code scanning workflow languages to include java. If you have any feedback or questions, please use this discussion thread or open an issue if you encounter any problems.

Kotlin support is an extension of our existing Java support, and benefits from all of our existing CodeQL queries for Java, for both mobile and server-side applications. We’ve also improved and added a range of mobile-specific queries, covering issues such as handling of Intents, Webview validation problems, fragment injection and more.

CodeQL support for Kotlin has already been used to identify novel real-world vulnerabilities in popular apps, from task management to productivity platforms. You can watch the GitHub Universe talk on how CodeQL was used to identify vulnerabilities like these here.

Kotlin beta support is available by default in GitHub.com code scanning, the CodeQL CLI, and the CodeQL extension for VS Code. GitHub Enterprise Server (GHES) version 3.8 will include this beta release.

See more

CodeQL code scanning launches Ruby analysis support in GA

Last year, we launched Ruby analysis support in beta for GitHub code scanning. Today, we're announcing the general availability of this feature — covering even more vulnerabilities in Ruby code.

Ruby is part of the top 10 most popular languages on GitHub today. In the past year alone, code scanning (powered by the CodeQL engine) helped Ruby developers resolve more than 4,000 security issues. Set up code scanning on your repositories today and receive actionable security alerts right on your pull-requests.

Since shipping in beta, our Ruby analysis has more than doubled the number of common weaknesses (CWEs) that it can detect. A total of 30 rules check your code for a range of vulnerabilities, including cross-site scripting (XSS), regular expression denial-of-service (ReDoS), SQL injection, and more. Additional library and framework coverage for Ruby-on-Rails ensures that web service developers get even more precise results. We currently support all common Ruby versions, up to and including 3.1. Check out the documentation for more details on compatibility.

Ruby support is available by default in GitHub.com code scanning, the CodeQL CLI, and the CodeQL extension for VS Code. GitHub Enterprise Server (GHES) version 3.4 shipped with Ruby (beta) support, and GHES 3.8 will include this GA release.

See more

Code scanning now supports using CodeQL packs on GitHub.com and GHES

CodeQL comes with a built-in package manager that helps you share and manage custom queries. Last year, we announced the public beta of CodeQL packaging — including direct integration into GitHub code scanning. This makes it easier to roll out custom queries to your repositories and gives you full control over exactly which queries are run.

This functionality will soon be released for users of GitHub Enterprise Server (GHES): it will be available with GHES 3.7. This release also includes support for using packs that are published to GitHub Container Registries (GHCR) hosted on GHES.

How do I use CodeQL query packs in code scanning?

To use CodeQL query packs in code scanning, specify a with: packs: entry in the uses: github/codeql-action/init@v2 section of your code scanning workflow. By default code scanning downloads packs the from GHCR on GitHub.com, so if you want to run one of the standard CodeQL query packs or any other public CodeQL query pack, then simply include the pack scope/name and version requirements here. You can find the full documentation here.

If you want to run packs from the GHCR on GHES, then you need to tell code scanning how to access and authenticate to the appropriate registry. For an example of how to do this in your code scanning workflow, see Downloading CodeQL packs from GitHub Enterprise Server in the GitHub documentation.

How do I publish my own CodeQL packs?

You can publish you own CodeQL packs using the CodeQL CLI. By default, the CodeQL CLI publishes packs to the GHCR on GitHub.com. If you want to publish packs to the GHCR associated with your instance of GHES, you need to tell the CodeQL CLI how to access and authenticate to the registry you want to work with. For a full example of how to specify these details, see Working with CodeQL packs on GitHub Enterprise Server in the CodeQL CLI documentation.

Where can I find more information about CodeQL packaging and code scanning?

This changelog post only provides a brief summary of how you can use CodeQL packs in code scanning. For more information, see:

See more

CodeQL code scanning now supports customizing build configurations for Go analysis

CodeQL now officially supports customizing the build configuration for Go analysis in the Actions workflow file. This aligns the Go configuration experience with the C/C++, C#, and Java analysis. The new customization options allow for more flexibility, for example when the build fails, or if analysis is desired on different source files.

All your existing CodeQL workflows for Go analysis will continue to work and continue to be supported. You don’t need to take any action to keep Go analysis running.

Example Actions workflow steps using Go build customization

steps:
  - name: Checkout repository
    uses: actions/checkout@v3

  - name: Initialize CodeQL
    uses: github/codeql-action/init@v2
    with:
      languages: go

  - name: Build code
    run:
      # You can modify these commands or add new commands to customize the build process
      make bootstrap
      make release

  - name: Perform CodeQL Analysis
    uses: github/codeql-action/analyze@v2

Learn more about CodeQL and code scanning.

See more

CodeQL for VS Code: download CodeQL databases from GitHub.com

We have started creating and storing CodeQL databases for the most popular open-source projects on GitHub.com. If you use CodeQL for security research, you can now obtain these databases easily and directly through the CodeQL extension for Visual Studio Code, which makes it much easier to write and run your own custom CodeQL queries.

Using CodeQL for security research

The CodeQL engine powers GitHub code scanning: it analyses source code and flags up potential security problems (for example, in pull requests). By default, code scanning runs a large set of open source queries that are able to identify the most important and common security problems.

CodeQL is also a powerful tool for variant analysis and other types of security research. CodeQL treats source code as data, and anyone can write custom CodeQL queries to explore a codebase and identify vulnerabilities. Like code search on steroids!

The first step of any CodeQL analysis is extracting the source code into a CodeQL database. This database contains a relational representation of the source code — including elements like the abstract syntax tree, the data flow graph, and the control flow graph. You can create CodeQL databases yourself using the CodeQL CLI, but with the feature we shipped today, it's much quicker to get started: you can download a ready-built CodeQL database from GitHub.com.

Downloading CodeQL databases from GitHub.com in VS Code

To download a CodeQL database for use in the CodeQL extension in VS Code:

  1. Make sure you have set up the CodeQL extension for VS Code. For more information, see Setting up CodeQL in Visual Studio Code.
  2. Open the CodeQL databases view in the extension.
  3. Hover over the sidebar, click the GitHub icon, and specify the owner/repo identifier of the public repository you'd like to analyze.

    image

Once you've downloaded a CodeQL database, you're ready to start your research. Find more information in the CodeQL documentation.

FAQs

How many CodeQL databases are available?

We currently store databases for over 200,000 repositories on GitHub.com. That list is constantly growing and evolving to make sure that it includes the most interesting codebases for security research.

What languages are can you download CodeQL databases for?

We create and store databases for all of the languages that we support in CodeQL code scanning. For more information, see About code scanning with CodeQL.

Can I download CodeQL databases outside VS Code?

Yes, you can also download CodeQL databases using the GitHub REST API. For more information, see Downloading databases from GitHub.com in the CodeQL CLI documentation.

Why is there no CodeQL codebase available for my favourite open source repository?

If there is a repository that you'd like to analyze, but a CodeQL database is not available yet, then you can trigger the creation (and storing) of a database by enabling GitHub code scanning with the CodeQL engine. Alternatively, you could fork the repository and enable code scanning on the fork. For more information, see the code scanning documentation.

See more

Debugging CodeQL analysis in code scanning made easier by obtaining detailed logs and debugging artifacts from the CodeQL Action

It's now easier to debug CodeQL analysis problems in code scanning: click Re-run jobs from the GitHub Actions workflow run page, check the Enable debug logging box, and hit the Re-run jobs button.

Re-run all jobs

The data will be uploaded as an Actions artifact named debug-artifacts, attached to the workflow run. Such artifacts contain CodeQL logs, CodeQL databases, and the SARIF files that were produced.

Actions artifacts

These artifacts will help you when you're debugging problems with CodeQL code scanning. When contacting GitHub support, you might be asked for this data.

As part of the analysis, CodeQL extracts your source code into a relational database format. The debug artifacts include more detailed information about CodeQL extraction errors and warnings that occurred during database creation. If you want to permanently enable debug logging for the CodeQL analysis, or would like more information about troubleshooting CodeQL, please follow these instructions.

This feature is now available to all users on GitHub.com and will also be available in GitHub Enterprise Server 3.7.

See more

Code scanning: deprecation of CodeQL Action v1

On March 30, 2022, we released CodeQL Action v2, which runs on the Node.js 16 runtime. The CodeQL Action v1 will be deprecated at the same time as GHES 3.3, which is currently scheduled for January 2023.

How does this affect me?

Users of GitHub.com, GitHub AE, and GitHub Enterprise Server 3.5 (and later)

All users of GitHub code scanning (which by default uses the CodeQL analysis engine) on GitHub Actions on the following platforms should update their workflow files:

  • GitHub.com (including open source repositories, users of GitHub Teams and GitHub Enterprise Cloud)
  • GitHub AE
  • GitHub Enterprise Server (GHES) 3.5 and later

Users of the above-mentioned platforms should update their CodeQL workflow file(s) to refer to the new v2 version of the CodeQL Action.

Users of GitHub Enterprise Server 3.4 (and older)

Users of GitHub Enterprise Server 3.4 (and older) are not recommended to update their configuration to use the v2 version of the CodeQL Action:

  • GHES 3.3 (and older) does not support running Actions using the Node 16 runtime and is therefore unable to run the v2 version of the CodeQL Action. Please upgrade to a newer version of GitHub Enterprise Server prior to changing your CodeQL Action workflow files.
  • While GHES 3.4 does support Node 16 Actions, it does not ship with v2 of the CodeQL Action. Users who want to migrate to v2 on GHES 3.4 should request that their system administrator enables GitHub Connect to download v2 onto GHES before updating their workflow files.

The upcoming release of GitHub Enterprise Server 3.5 will ship with v2 of the CodeQL Action included.

Exactly what do I need to change?

To upgrade to the CodeQL Action v2, open your CodeQL workflow file(s) in the .github directory of your repository and look for references to:

  • github/codeql-action/init@v1
  • github/codeql-action/autobuild@v1
  • github/codeql-action/analyze@v1
  • github/codeql-action/upload-sarif@v1

These entries need to be replaced with their v2 equivalents:

  • github/codeql-action/init@v2
  • github/codeql-action/autobuild@v2
  • github/codeql-action/analyze@v2
  • github/codeql-action/upload-sarif@v2

Can I use Dependabot to help me with this upgrade?

Yes, you can! For more details on how to configure Dependabot to automatically upgrade your Actions dependencies, please see this page.

What happens in January 2023?

In January 2023, the CodeQL Action v1 will be officially deprecated (at the same time as the GHES 3.3 deprecation). At that point, no new updates will be made to v1, which means that new CodeQL analysis capabilities will only be available to users of v2. We will keep a close eye on the migration progress across GitHub. If many workflow files still refer to v1 of the CodeQL Action we might consider scheduling one or more brownout moments later in the year to increase awareness.

See more
View more changes