- Personal access tokens
- OAuth 2.0 tokens
- Impersonation tokens
- Project access tokens
- Group access tokens
- Deploy tokens
- Deploy keys
- Runner authentication tokens
- Runner registration tokens (deprecated)
- CI/CD job tokens
- GitLab cluster agent tokens
- Other tokens
- Available scopes
- Token prefixes
- Security considerations
GitLab token overview
This document lists tokens used in GitLab, their purpose and, where applicable, security guidance.
Personal access tokens
You can create personal access tokens to authenticate with:
- The GitLab API.
- GitLab repositories.
- The GitLab registry.
You can limit the scope and expiration date of your personal access tokens. By default, they inherit permissions from the user who created them.
You can use the personal access tokens API to programmatically take action, such as rotating a personal access token.
You receive an email when your personal access tokens are expiring soon.
OAuth 2.0 tokens
GitLab can serve as an OAuth 2.0 provider to allow other services to access the GitLab API on a user’s behalf.
You can limit the scope and lifetime of your OAuth 2.0 tokens.
Impersonation tokens
An impersonation token is a special type of personal access token. It can be created only by an administrator for a specific user. Impersonation tokens can help you build applications or scripts that authenticate with the GitLab API, repositories, and the GitLab registry as a specific user.
You can limit the scope and set an expiration date for an impersonation token.
Project access tokens
Project access tokens are scoped to a project. Like personal access tokens, you can use them to authenticate with:
- The GitLab API.
- GitLab repositories.
- The GitLab registry.
You can limit the scope and expiration date of project access tokens. When you create a project access token, GitLab creates a bot user for projects. Bot users for projects are service accounts and do not count as licensed seats.
You can use the project access tokens API to programmatically take action, such as rotating a project access token.
Members of a project with at least the Maintainer role receive an email when project access tokens are nearly expired.
Group access tokens
Group access tokens are scoped to a group. Like personal access tokens, you can use them to authenticate with:
- The GitLab API.
- GitLab repositories.
- The GitLab registry.
You can limit the scope and expiration date of group access tokens. When you create a group access token, GitLab creates a bot user for groups. Bot users for groups are service accounts and do not count as licensed seats.
You can use the group access tokens API to programmatically take action, such as rotating a group access token.
Members of a group with the Owner role receive an email when group access tokens are nearly expired.
Deploy tokens
Deploy tokens allow you to clone, push, and pull packages and container registry images of a project without a user and a password. Deploy tokens cannot be used with the GitLab API.
To manage deploy tokens, you must be a member of a project with at least the Maintainer role.
Deploy keys
Deploy keys allow read-only or read-write access to your repositories by importing an SSH public key into your GitLab instance. Deploy keys cannot be used with the GitLab API or the registry.
You can use deploy keys to clone repositories to your continuous integration server without setting up a fake user account.
To add or enable a deploy key for a project, you must have at least the Maintainer role.
Runner authentication tokens
In GitLab 16.0 and later, to register a runner, you can use a runner authentication token instead of a runner registration token. Runner registration tokens are deprecated.
After you create a runner and its configuration, you receive a runner authentication token
that you use to register the runner. The runner authentication token is stored locally in
the config.toml
file,
which you use to configure the runner.
The runner uses the runner authentication token to authenticate with GitLab when it picks up jobs from the job queue. After the runner authenticates with GitLab, the runner receives a job token, which it uses to execute the job.
The runner authentication token stays on the runner machine. The execution environments for the following executors have access to only the job token and not the runner authentication token:
- Docker Machine
- Kubernetes
- VirtualBox
- Parallels
- SSH
Malicious access to a runner’s file system might expose the
config.toml
file and the runner authentication token. The attacker
could use the runner authentication token to
clone the runner.
You can use the runners API to rotate or revoke a runner authentication token.
Runner registration tokens (deprecated)
Runner registration tokens are used to register a runner with GitLab. Group or project owners or instance administrators can obtain them through the GitLab user interface. The registration token is limited to runner registration and has no further scope.
You can use the runner registration token to add runners that execute jobs in a project or group. The runner has access to the project’s code, so be careful when assigning permissions to projects or groups.
CI/CD job tokens
The CI/CD job token is a short-lived token valid only for the duration of a job. It gives a CI/CD job access to a limited number of API endpoints. API authentication uses the job token by using the authorization of the user triggering the job.
The job token is secured by its short lifetime and limited scope. This token could be leaked if
multiple jobs run on the same machine (for example, with the shell runner). You can use the project allow list to further limit what the job token can access.
On Docker Machine runners, you should configure
MaxBuilds=1
to ensure runner machines run only one build
and are destroyed afterwards. Provisioning takes time,
so this configuration can affect performance.
GitLab cluster agent tokens
When you register a GitLab agent for Kubernetes, GitLab generates an access token to authenticate the cluster agent with GitLab.
To revoke this cluster agent token, you can either:
- Revoke the token with the agents API.
- Reset the token.
For both methods, you must know the token, agent, and project IDs. To find this information, use the Rails console:
# Find token ID
Clusters::AgentToken.find_by_token('glagent-xxx').id
# Find agent ID
Clusters::AgentToken.find_by_token('glagent-xxx').agent.id
=> 1234
# Find project ID
Clusters::AgentToken.find_by_token('glagent-xxx').agent.project_id
=> 12345
You can also revoke a token directly in the Rails console:
# Revoke token with RevokeService, including generating an audit event
Clusters::AgentTokens::RevokeService.new(token: Clusters::AgentToken.find_by_token('glagent-xxx'), current_user: User.find_by_username('admin-user')).execute
# Revoke token manually, which does not generate an audit event
Clusters::AgentToken.find_by_token('glagent-xxx').revoke!
Other tokens
Feed token
Each user has a long-lived feed token that does not expire. Use this token to authenticate with:
- RSS readers, to load a personalized RSS feed.
- Calendar applications, to load a personalized calendar.
You cannot use this token to access any other data.
You can use the user-scoped feed token for all feeds. However, feed and calendar URLs are generated with a different token valid for only one feed.
Anyone who has your token can view your feed activity, including confidential issues, as if they were you. If you think your token has leaked, reset the token immediately.
Disable a feed token
Prerequisites:
- You must be an administrator.
- On the left sidebar, at the bottom, select Admin.
- Select Settings > General.
- Expand Visibility and access controls.
- Under Feed token, select the Disable feed token checkbox, then select Save changes.
Incoming email token
Each user has an incoming email token that does not expire. The token is included in email addresses associated with a personal project. You use this token to create a new issue by email.
You cannot use this token to access any other data. Anyone who has your token can create issues and merge requests as if they were you. If you think your token has leaked, reset the token immediately.
Available scopes
This table shows default scopes per token. For some tokens, you can limit scopes further when you create the token.
Token name | API access | Registry access | Repository access |
---|---|---|---|
Personal access token | Yes | Yes | Yes |
OAuth 2.0 token | Yes | No | Yes |
Impersonation token | Yes | Yes | Yes |
Project access token | Yes1 | Yes1 | Yes1 |
Group access token | Yes2 | Yes2 | Yes2 |
Deploy token | No | Yes | Yes |
Deploy key | No | No | Yes |
Runner registration token | No | No | Limited3 |
Runner authentication token | No | No | Limited3 |
Job token | Limited4 | No | Yes |
Footnotes:
- Limited to the one project.
- Limited to the one group.
- Runner registration and authentication tokens don’t provide direct access to repositories, but can be used to register and authenticate new runners that can execute jobs which do have access to repositories.
- Only certain endpoints.
Token prefixes
The following table shows the prefixes for each type of token.
Token name | Prefix |
---|---|
Personal access token |
glpat-
|
OAuth Application Secret |
gloas-
|
Impersonation token |
glpat-
|
Project access token |
glpat-
|
Group access token |
glpat-
|
Deploy token |
gldt- (Added in GitLab 16.7)
|
Runner authentication token |
glrt-
|
CI/CD Job token |
glcbt- • (Introduced in GitLab 16.8 behind a feature flag named prefix_ci_build_tokens . Disabled by default.) • (Generally available in GitLab 16.9. Feature flag prefix_ci_build_tokens removed.)
|
Trigger token |
glptt-
|
Feed token |
glft-
|
Incoming mail token |
glimt-
|
GitLab agent for Kubernetes token |
glagent-
|
GitLab session cookies |
_gitlab_session=
|
SCIM Tokens |
glsoat- • (Introduced in GitLab 16.8 behind a feature flag named prefix_scim_tokens . Disabled by default.) • (Generally available in GitLab 16.9. Feature flag prefix_scim_tokens removed.)
|
Feature Flags Client token |
glffct-
|
Security considerations
To keep your tokens secure:
- Treat access tokens like passwords and keep them secure.
- When creating a scoped token, consider using the most limited scope possible to reduce the impact of accidentally leaking the token.
- When creating a token, consider setting a token that expires when your task is complete. For example, if you need to perform a one-time import, set the token to expire after a few hours.
- If you set up a demo environment to showcase a project you have been working on, and you record a video or write a blog post describing that project, make sure you don’t accidentally leak a secret. After the demo is finished, revoke all the secrets created during the demo.
- Adding access tokens to URLs is a security risk, especially when cloning or adding a remote, because Git writes URLs to its
.git/config
file in plaintext. URLs are also often logged by proxies and application servers, which leaks those credentials to system administrators. Instead, pass an access token to an API call with a header likePrivate-Token
. - You can store tokens using Git credential storage.
- Review all active access tokens of all types on a regular basis and revoke any you don’t need.
Do not:
- Store tokens in plaintext in your projects. If the token is an external secret for GitLab CI/CD, review how to use external secrets in CI recommendations.
- Include tokens when pasting code, console commands, or log outputs into an issue, MR description, or comment.
- Log credentials in the console logs or artifacts. Consider protecting and masking your credentials.