You can see examples of pre-receive hooks for GitHub Enterprise Server in the
Writing a pre-receive hook script
A pre-receive hook script executes in a pre-receive hook environment on your GitHub Enterprise Server instance. When you create a pre-receive hook script, consider the available input, output, exit status, and environment variables.
After a push occurs and before any refs are updated for the remote repository, the
git-receive-pack process on your GitHub Enterprise Server instance invokes the pre-receive hook script. Standard input for the script,
stdin, is a string containing a line for each ref to update. Each line contains the old object name for the ref, the new object name for the ref, and the full name of the ref.
<old-value> SP <new-value> SP <ref-name> LF
This string represents the following arguments.
|Old object name stored in the ref.|
When you create a new ref, the value is 40 zeroes.
|New object name to be stored in the ref.|
When you delete a ref, the value is 40 zeroes.
|The full name of the ref.|
For more information about
git-receive-pack, see "git-receive-pack" in the Git documentation. For more information about refs, see "Git References" in Pro Git.
The standard output for the script,
stdout, is passed back to the client. Any
echo statements will be visible to the user on the command line or in the user interface.
The exit status of a pre-receive script determines if the push will be accepted.
|0||The push will be accepted.|
|non-zero||The push will be rejected.|
In addition to the standard input for your pre-receive hook script,
stdin, GitHub Enterprise Server makes the following variables available in the Bash environment for your script's execution. For more information about
stdin for your pre-receive hook script, see "Input (
Different environment variables are available to your pre-receive hook script depending on what triggers the script to run.
- Always available
- Available for pushes from the web interface or API
- Available for pull request merges
- Available for pushes using SSH authentication
The following variables are always available in the pre-receive hook environment.
|Path to the remote repository on the instance||/data/user/repositories/a/ab/|
|The number of push options that were sent by the client with ||1|
|Where N is an integer starting at 0, this variable contains the push option string that was sent by the client. The first option that was sent is stored in ||abcd|
|User-agent string sent by the Git client that pushed the changes||git/2.0.0|
|Name of the repository being updated in NAME/OWNER format||octo-org/hello-enterprise|
|Boolean representing whether the repository being updated is public|
|IP address of client that initiated the push||192.0.2.1|
|Username for account that initiated the push||octocat|
Available for pushes from the web interface or API
$GITHUB_VIA variable is available in the pre-receive hook environment when the ref update that triggers the hook occurs via either the web interface or the API for GitHub Enterprise Server. The value describes the action that updated the ref.
auto-merge deployment api
|Automatic merge of the base branch via a deployment created with the API||"Create a deployment" in the REST API documentation|
|Change to a file's contents in the web interface||"Editing files"|
branch merge api
|Merge of a branch via the API||"Merge a branch" in the REST API documentation|
branches page delete button
|Deletion of a branch in the web interface||"Creating and deleting branches within your repository"|
git refs create api
|Creation of a ref via the API||"Git database" in the REST API documentation|
git refs delete api
|Deletion of a ref via the API||"Git database" in the REST API documentation|
git refs update api
|Update of a ref via the API||"Git database" in the REST API documentation|
git repo contents api
|Change to a file's contents via the API||"Create or update file contents" in the REST API documentation|
|Merge of a pull request using auto-merge||"Automatically merging a pull request"|
merge base into head
|Update of the topic branch from the base branch when the base branch requires strict status checks (via Update branch in a pull request, for example)||"About protected branches"|
pull request branch delete button
|Deletion of a topic branch from a pull request in the web interface||"Deleting and restoring branches in a pull request"|
pull request branch undo button
|Restoration of a topic branch from a pull request in the web interface||"Deleting and restoring branches in a pull request"|
pull request merge api
|Merge of a pull request via the API||"Pulls" in the REST API documentation|
pull request merge button
|Merge of a pull request in the web interface||"Merging a pull request"|
pull request revert button
|Revert of a pull request||"Reverting a pull request"|
releases delete button
|Deletion of a release||"Managing releases in a repository"|
stafftools branch restore
|Restoration of a branch from the site admin dashboard||"Site admin dashboard"|
tag create api
|Creation of a tag via the API||"Git database" in the REST API documentation|
|Commit via Subversion||"Support for Subversion clients"|
web branch create
|Creation of a branch via the web interface||"Creating and deleting branches within your repository"|
Available for pull request merges
The following variables are available in the pre-receive hook environment when the push that triggers the hook is a push due to the merge of a pull request.
|Username of account that authored the pull request||octocat|
|The name of the pull request's topic branch, in the format |
|The name of the pull request's base branch, in the format ||octocat:main|
Available for pushes using SSH authentication
|The public key fingerprint for the user who pushed the changes||a1:b2:c3:d4:e5:f6:g7:h8:i9:j0:k1:l2:m3:n4:o5:p6|
Setting permissions and pushing a pre-receive hook to GitHub Enterprise Server
A pre-receive hook script is contained in a repository on your GitHub Enterprise Server instance. A site administrator must take into consideration the repository permissions and ensure that only the appropriate users have access.
We recommend consolidating hooks to a single repository. If the consolidated hook repository is public, the
README.md can be used to explain policy enforcements. Also, contributions can be accepted via pull requests. However, pre-receive hooks can only be added from the default branch. For a testing workflow, forks of the repository with configuration should be used.
For Mac users, ensure the scripts have execute permissions:
$ sudo chmod +x SCRIPT_FILE.sh
For Windows users, ensure the scripts have execute permissions:
git update-index --chmod=+x SCRIPT_FILE.sh
Commit and push to the designated repository for pre-receive hooks on your GitHub Enterprise Server instance.
$ git commit -m "YOUR COMMIT MESSAGE" $ git push
Create the pre-receive hook on the GitHub Enterprise Server instance.
Testing pre-receive scripts locally
You can test a pre-receive hook script locally before you create or update it on your GitHub Enterprise Server instance. One method is to create a local Docker environment to act as a remote repository that can execute the pre-receive hook.
Ensure Docker is installed locally.
Create a file called
FROM gliderlabs/alpine:3.3 RUN \ apk add --no-cache git openssh bash && \ ssh-keygen -A && \ sed -i "s/#AuthorizedKeysFile/AuthorizedKeysFile/g" /etc/ssh/sshd_config && \ adduser git -D -G root -h /home/git -s /bin/bash && \ passwd -d git && \ su git -c "mkdir /home/git/.ssh && \ ssh-keygen -t ed25519 -f /home/git/.ssh/id_ed25519 -P '' && \ mv /home/git/.ssh/id_ed25519.pub /home/git/.ssh/authorized_keys && \ mkdir /home/git/test.git && \ git --bare init /home/git/test.git" VOLUME ["/home/git/.ssh", "/home/git/test.git/hooks"] WORKDIR /home/git CMD ["/usr/sbin/sshd", "-D"]
Create a test pre-receive script called
always_reject.sh. This example script will reject all pushes, which is useful for locking a repository:
#!/usr/bin/env bash echo "error: rejecting all pushes" exit 1
always_reject.shscripts has execute permissions:
$ chmod +x always_reject.sh
From the directory containing
Dockerfile.dev, build an image:
$ docker build -f Dockerfile.dev -t pre-receive.dev . > Sending build context to Docker daemon 3.584 kB > Step 1 : FROM gliderlabs/alpine:3.3 > ---> 8944964f99f4 > Step 2 : RUN apk add --no-cache git openssh bash && ssh-keygen -A && sed -i "s/#AuthorizedKeysFile/AuthorizedKeysFile/g" /etc/ssh/sshd_config && adduser git -D -G root -h /home/git -s /bin/bash && passwd -d git && su git -c "mkdir /home/git/.ssh && ssh-keygen -t ed25519 -f /home/git/.ssh/id_ed25519 -P ' && mv /home/git/.ssh/id_ed25519.pub /home/git/.ssh/authorized_keys && mkdir /home/git/test.git && git --bare init /home/git/test.git" > ---> Running in e9d79ab3b92c > fetch http://alpine.gliderlabs.com/alpine/v3.3/main/x86_64/APKINDEX.tar.gz > fetch http://alpine.gliderlabs.com/alpine/v3.3/community/x86_64/APKINDEX.tar.gz ....truncated output.... > OK: 34 MiB in 26 packages > ssh-keygen: generating new host keys: RSA DSA ECDSA ED25519 > Password for git changed by root > Generating public/private ed25519 key pair. > Your identification has been saved in /home/git/.ssh/id_ed25519. > Your public key has been saved in /home/git/.ssh/id_ed25519.pub. ....truncated output.... > Initialized empty Git repository in /home/git/test.git/ > Successfully built dd8610c24f82
Run a data container that contains a generated SSH key:
$ docker run --name data pre-receive.dev /bin/true
Copy the test pre-receive hook
always_reject.shinto the data container:
$ docker cp always_reject.sh data:/home/git/test.git/hooks/pre-receive
Run an application container that runs
sshdand executes the hook. Take note of the container id that is returned:
$ docker run -d -p 52311:22 --volumes-from data pre-receive.dev > 7f888bc700b8d23405dbcaf039e6c71d486793cad7d8ae4dd184f4a47000bc58
Copy the generated SSH key from the data container to the local machine:
$ docker cp data:/home/git/.ssh/id_ed25519 .
Modify the remote of a test repository and push to the
test.gitrepo within the Docker container. This example uses
email@example.com:octocat/Hello-World.gitbut you can use any repository you want. This example assumes your local machine (127.0.0.1) is binding port 52311, but you can use a different IP address if docker is running on a remote machine.
$ git clone firstname.lastname@example.org:octocat/Hello-World.git $ cd Hello-World $ git remote add test email@example.com:test.git $ GIT_SSH_COMMAND="ssh -o UserKnownHostsFile=/dev/null -o StrictHostKeyChecking=no -p 52311 -i ../id_ed25519" git push -u test main > Warning: Permanently added '[192.168.99.100]:52311' (ECDSA) to the list of known hosts. > Counting objects: 7, done. > Delta compression using up to 4 threads. > Compressing objects: 100% (3/3), done. > Writing objects: 100% (7/7), 700 bytes | 0 bytes/s, done. > Total 7 (delta 0), reused 7 (delta 0) > remote: error: rejecting all pushes > To firstname.lastname@example.org:test.git > ! [remote rejected] main -> main (pre-receive hook declined) > error: failed to push some refs to 'email@example.com:test.git'
Notice that the push was rejected after executing the pre-receive hook and echoing the output from the script.
- "Customizing Git - An Example Git-Enforced Policy" from the Pro Git website