Where
-Infinity
0
Severity
9.8
Integer Overflow
AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H

cmark-gfm is GitHub's extended version of the C reference implementation of CommonMark. Prior to versions 0.29.0.gfm.3 and 0.28.3.gfm.21, an integer overflow in cmark-gfm's table row parsing table.c:rowfromstring may lead to heap memory corruption when parsing tables who's marker rows contain more than UINT16MAX columns. The impact of this heap corruption ranges from Information Leak to Arbitrary Code Execution depending on how and where cmark-gfm is used. If cmark-gfm is used for rendering remote user controlled markdown, this vulnerability may lead to Remote Code Execution (RCE) in applications employing affected versions of the cmark-gfm library. This vulnerability has been patched in the following cmark-gfm versions 0.29.0.gfm.3 and 0.28.3.gfm.21. A workaround is available. The vulnerability exists in the table markdown extensions of cmark-gfm. Disabling the table extension will prevent this vulnerability from being triggered.

1 / 2
First published (updated )
Severity
7.5
AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H

cmark-gfm is GitHub's fork of cmark, a CommonMark parsing and rendering library and program in C. In versions prior to 0.29.0.gfm.6 a polynomial time complexity issue in cmark-gfm's autolink extension may lead to unbounded resource exhaustion and subsequent denial of service. Users may verify the patch by running python3 -c 'print("![l" 100000 + "\n")' | ./cmark-gfm -e autolink, which will resource exhaust on unpatched cmark-gfm but render correctly on patched cmark-gfm. This vulnerability has been patched in 0.29.0.gfm.6. Users are advised to upgrade. Users unable to upgrade should disable the use of the autolink extension.

1 / 2
Source: MITRE
First published (updated )
Severity
8.3
XSS
AV:N/AC:L/AT:N/PR:H/UI:A/VC:H/VI:H/VA:N/SC:N/SI:L/SA:N/E:X/CR:X/IR:X/AR:X/MAV:X/MAC:X/MAT:X/MPR:X/MUI:X/MVC:X/MVI:X/MVA:X/MSC:X/MSI:X/MSA:X/S:X/AU:X/R:X/V:X/RE:X/U:X

A stored cross-site scripting vulnerability was identified in GitHub Enterprise Server that allowed an organization member with permission to manage custom-property definitions and edit property values to add an attacker-controlled SSH authentication key to the personal account of another member through HTML stored in a single-select custom-property option. Exploitation required another definition manager in the same organization to attempt to remove the poisoned option while it was in use and save the definition with a valid sudo session. Repository access through the added key depended on the permissions of the victim and independent authorization controls. This vulnerability affected supported GitHub Enterprise Server prior to 3.23 and was fixed in versions 3.19.16, 3.20.12, 3.21.10, and 3.22.5. This vulnerability was reported via the GitHub Bug Bounty program.

First published (updated )
Severity
7.7
SSRF
AV:N/AC:L/AT:P/PR:L/UI:N/VC:H/VI:H/VA:H/SC:N/SI:N/SA:N/E:X/CR:X/IR:X/AR:X/MAV:X/MAC:X/MAT:X/MPR:X/MUI:X/MVC:X/MVI:X/MVA:X/MSC:X/MSI:X/MSA:X/S:X/AU:X/R:X/V:X/RE:X/U:X

A server-side request forgery vulnerability in GitHub Enterprise Server allowed an authenticated user with repository write access to send requests to internal services through Slack webhook validity checks, potentially leading to remote code execution on the appliance. Exploitation required Secret Protection with secret scanning and validity checks for partner patterns enabled. This vulnerability affected all versions prior to 3.23 and was fixed in versions 3.22.5, 3.21.10, and 3.20.12. This vulnerability was reported via the GitHub Bug Bounty program.

First published (updated )
Severity
5.3
CRLF Injection
AV:N/AC:L/AT:N/PR:L/UI:N/VC:N/VI:L/VA:N/SC:N/SI:N/SA:N/E:X/CR:X/IR:X/AR:X/MAV:X/MAC:X/MAT:X/MPR:X/MUI:X/MVC:X/MVI:X/MVA:X/MSC:X/MSI:X/MSA:X/S:X/AU:X/R:X/V:X/RE:X/U:X

A CRLF injection vulnerability was identified in GitHub Enterprise Server that allowed a user with write access to a repository to bypass branch protections and update branches or tags in that repository through specially crafted input when resolving pull request conflicts. This vulnerability affected all versions of GitHub Enterprise Server prior to 3.23 and was fixed in versions 3.19.16, 3.20.12, 3.21.10, and 3.22.5. This vulnerability was reported via the GitHub Bug Bounty program.

First published (updated )
Severity
8.6
AV:N/AC:L/AT:N/PR:H/UI:N/VC:H/VI:H/VA:H/SC:N/SI:N/SA:N/E:X/CR:X/IR:X/AR:X/MAV:X/MAC:X/MAT:X/MPR:X/MUI:X/MVC:X/MVI:X/MVA:X/MSC:X/MSI:X/MSA:X/S:X/AU:X/R:X/V:X/RE:X/U:X

A site administrator could modify files on the appliance host with root privileges and disclose private repository data due to improper path validation in GitHub Enterprise Server. Exploitation required a crafted pre-receive hook environment. This vulnerability affected supported GitHub Enterprise Server versions prior to 3.23 and was fixed in versions 3.19.16, 3.20.12, 3.21.10, and 3.22.5. This vulnerability was reported via the GitHub Bug Bounty program.

First published (updated )
Severity
8.6
AV:N/AC:L/AT:N/PR:L/UI:N/VC:H/VI:H/VA:N/SC:N/SI:N/SA:N/E:X/CR:X/IR:X/AR:X/MAV:X/MAC:X/MAT:X/MPR:X/MUI:X/MVC:X/MVI:X/MVA:X/MSC:X/MSI:X/MSA:X/S:X/AU:X/R:X/V:X/RE:X/U:X

An SSH certificate authentication bypass vulnerability in GitHub Enterprise Server allowed an authenticated organization member to read and modify another user's private repositories within the trusted certificate authority's scope. Exploitation required an instance using built-in authentication and a valid certificate issued to the attacker by a trusted SSH certificate authority. This vulnerability affected supported GitHub Enterprise Server releases prior to 3.23 and was fixed in versions 3.19.16, 3.20.12, 3.21.10, and 3.22.5. This vulnerability was reported via the GitHub Bug Bounty program.

First published (updated )
Severity
7.4
AV:N/AC:L/AT:P/PR:H/UI:N/VC:H/VI:H/VA:N/SC:N/SI:N/SA:N/E:X/CR:X/IR:X/AR:X/MAV:X/MAC:X/MAT:X/MPR:X/MUI:X/MVC:X/MVI:X/MVA:X/MSC:X/MSI:X/MSA:X/S:X/AU:X/R:X/V:X/RE:X/U:X

An incorrect authorization vulnerability was identified in GitHub Enterprise Server that allowed an authenticated user whose numeric user ID matched an enterprise's numeric ID to manage that enterprise's GitHub Apps and generate private keys without enterprise membership. GitHub App ownership checks compared numeric IDs without checking the owner's account type. Exploitation required control of a session for the colliding user account and an enterprise-owned GitHub App. A generated private key could be used to authenticate as the App and access repositories where it was installed, within its granted permissions. On standard fresh installations, the documented collision involved the non-loginable built-in ghost account; the reported GHES reproduction used site administrator impersonation. This vulnerability affected supported GitHub Enterprise Server releases prior to 3.23 and was fixed in versions 3.19.16, 3.20.12, 3.21.10, and 3.22.5. This vulnerability was reported via the GitHub Bug Bounty program.

First published (updated )
Severity
7.4
OS Command Injection, Command Injection
AV:N/AC:L/AT:P/PR:H/UI:N/VC:H/VI:H/VA:N/SC:N/SI:N/SA:N/E:X/CR:X/IR:X/AR:X/MAV:X/MAC:X/MAT:X/MPR:X/MUI:X/MVC:X/MVI:X/MVA:X/MSC:X/MSI:X/MSA:X/S:X/AU:X/R:X/V:X/RE:X/U:X

A command injection vulnerability was identified in GitHub Enterprise Server that allowed an authenticated site administrator to execute arbitrary commands as the host git user on peer Git server nodes outside the pre-receive hook sandbox by supplying a crafted pre-receive environment download URL. Exploitation required a clustered deployment using custom pre-receive hook environments and at least one peer Git server node. This vulnerability affected all versions of GitHub Enterprise Server prior to 3.23 and was fixed in versions 3.22.5, 3.21.10, 3.20.12, and 3.19.16. This vulnerability was reported via the GitHub Bug Bounty program.

First published (updated )
Severity
7.1
AV:N/AC:L/AT:N/PR:L/UI:N/VC:H/VI:N/VA:N/SC:N/SI:N/SA:N/E:X/CR:X/IR:X/AR:X/MAV:X/MAC:X/MAT:X/MPR:X/MUI:X/MVC:X/MVI:X/MVA:X/MSC:X/MSI:X/MSA:X/S:X/AU:X/R:X/V:X/RE:X/U:X

An authorization bypass through user-controlled key vulnerability was identified in GitHub Enterprise Server that allowed authenticated users to read review-thread data from private repositories they could not access. An attacker required an authenticated account, a readable repository and pull request to use as the request context, and target review-thread identifiers. The thread-preview endpoint resolved supplied identifiers without restricting them to the authorized pull request, exposing private file paths and, when positioning and diff data were available, source-code excerpts, commit identifiers, and line metadata without victim interaction. This vulnerability affected supported GitHub Enterprise Server releases in the 3.21, 3.22 series and was fixed in versions 3.21.10, 3.22.5. This vulnerability was reported via the GitHub Bug Bounty program.

First published (updated )
Severity
3.3
AV:L/AC:L/PR:L/UI:N/S:U/C:L/I:N/A:N

Summary

A security vulnerability has been identified in GitHub CLI where gh auth status could potentially print a portion of the authentication token in plaintext, depending on the lexical structure of the token.

Affected users: - Authenticated users who ran gh auth status (without the --show-token flag) with a token type whose format contains an underscore after the prefix. This includes fine-grained personal access tokens (githubpat) and GitHub App installation/user access tokens (ghs, ghu, e.g., ghs<APPID><JWT>), including the Actions GITHUBTOKEN. Classic tokens such as gho and ghp have an underscore-free body and are not affected.

Details

The vulnerability stems from the way GitHub CLI masks the authentication token when reporting authentication state:

1. A user runs gh auth status without --show-token, expecting the token to be fully masked. 2. The masking logic redacts only the characters that follow the last underscore in the token, instead of redacting everything after the token prefix. 3. The partially masked token is written to the command output, where it can be captured or shared.

Because gh auth status is a diagnostic command commonly run interactively and inside automated environments, the leaked token characters can end up in CI/CD job logs, terminal scrollback, screen recordings, screenshots, or pasted output shared in bug reports and support requests.

In v2.97.0, GitHub CLI fully masks the token in gh auth status output, preserving only the known token prefix and redacting the remainder unless --show-token is explicitly provided.

Impact

Running gh auth status in a context where the output is captured, logged, or shared could leak a portion of an affected token. The affected token types are fine-grained personal access tokens (githubpat) and GitHub App installation/user access tokens (ghs, ghu, e.g., ghs<APPID><JWT>), including the Actions GITHUBTOKEN. Installation and Actions tokens are short-lived, which narrows the window in which a recovered token could be reused. See GitHub credential types.

Remediation and Mitigation

1. Upgrade gh to v2.97.0. 2. Rotate any affected token whose masked gh auth status output may have been captured or shared: - Fine-grained personal access tokens (githubpat): regenerate the token from your account settings. See Managing your personal access tokens. - GitHub App installation/user access tokens (ghs, ghu, e.g., ghs<APPID><JWT>): these are short-lived and rotate automatically (See GitHub credential types). If you believe a token was captured while still valid, you can revoke it immediately using the Revoke an installation access token REST endpoint (DELETE /installation/token). To cut off a third-party app's access more broadly, suspend or uninstall the app from the web UI. See Reviewing and modifying installed GitHub Apps. - Actions GITHUBTOKEN: this token is automatically invalidated when the job finishes, so no manual revocation is normally required. See GitHub credential types. 3. Until you can upgrade, avoid running gh auth status in contexts where the output is captured or shared, and manually redact the token line before sharing any output.

1 / 3
Source: GitHub
First published (updated )
Severity
3.3
AV:L/AC:L/PR:L/UI:N/S:U/C:L/I:N/A:N

GitHub CLI (gh) is GitHub's official command line tool. Prior to version 2.97.0, gh auth status masked only the characters after the last underscore in certain fine-grained personal access tokens and GitHub App tokens. As a result, part of an affected token could appear in terminal or CI output that is captured or shared. Authenticated users are affected if they ran gh auth status (without the --show-token flag) with a token type whose format contains an underscore after the prefix. This includes fine-grained personal access tokens (githubpat) and GitHub App installation and user access tokens (ghs, ghu; for example, ghs<APPID><JWT>), as well as the Actions GITHUBTOKEN. Classic tokens such as gho and ghp have an underscore-free body and are not affected. This issue is fixed in version 2.97.0.

First published (updated )
Severity
2.1
AV:N/AC:L/AT:P/PR:N/UI:A/VC:N/VI:N/VA:N/SC:N/SI:L/SA:N/E:X/CR:X/IR:X/AR:X/MAV:X/MAC:X/MAT:X/MPR:X/MUI:X/MVC:X/MVI:X/MVA:X/MSC:X/MSI:X/MSA:X/S:X/AU:X/R:X/V:X/RE:X/U:X

Summary

A security vulnerability has been identified in GitHub CLI that could allow attestation verification bypass when using gh attestation verify with --signer-repo or --signer-workflow flags. GitHub CLI builds the certificate matcher from these flag values without escaping regex metacharacters, so a repository or workflow name is treated as a regular expression rather than a literal string. GitHub permits characters such as  .  in organization, repository, and workflow path names, and  .  is a regex wildcard, so a malicious actor can create a lookalike name that satisfies a matcher intended for a different, trusted signer.

Details

GitHub CLI constructs a regular expression from the  --signer-repo  and  --signer-workflow  values and matches it against the signing certificate's SAN (Subject Alternative Name) during Sigstore attestation verification. The user-supplied value is interpolated into the pattern without escaping regex metacharacters, so any metacharacter permitted in a GitHub path is interpreted as regex syntax instead of a literal character.

The most practical case is  . , which GitHub allows in org/repo names and is a regex wildcard meaning "any single character." For example, an attacker-controlled repository named:

github/artifact.attestations-workflows

matches a verifier configured for the trusted signer:

github/artifact-attestations-workflows

because the unescaped  .  in the constructed pattern matches the  -  in the attacker's name (and any other single character at that position). This affected the --signer-repo and --signer-workflow flags but did not affect the --repo and --owner flags which include an exact string comparison.

In v2.97.0, the GitHub CLI escapes the user-supplied signer values before they are incorporated into the SAN matcher, so they are matched as literal strings.

Impact

The practical impact is limited, exploitation requires the attacker to create a plausible lookalike repository name and produce valid Sigstore attestations from it. This could undermine supply chain verification for CI/CD pipelines or policy gates that use these flags to pin trust to a specific signing workflow.

Remediation and Mitigation

1. Upgrade gh to v2.97.0 or later, which escapes signer values before matching. 2. As a workaround, use --repo or --owner flags instead of --signer-repo or --signer-workflow where applicable. 3. Manually inspect the signing certificate SAN in attestation output to confirm the signing workflow origin.

1 / 3
Source: GitHub
First published (updated )
Severity
5.1
Path Traversal
AV:N/AC:L/AT:N/PR:N/UI:A/VC:N/VI:N/VA:N/SC:L/SI:L/SA:N/E:X/CR:X/IR:X/AR:X/MAV:X/MAC:X/MAT:X/MPR:X/MUI:X/MVC:X/MVI:X/MVA:X/MSC:X/MSI:X/MSA:X/S:X/AU:X/R:X/V:X/RE:X/U:X

Summary

A path traversal vulnerability has been identified in the GitHub CLI where some HTTP request URLs, REST API calls in particular, were built without properly escaping their variable path components. A value containing URL path metacharacters can alter the request path and cause gh to address a different resource than intended.

Details

Not all request URLs in the codebase were built with proper escaping of their variable path components. When a component is interpolated into the path without percent encoding, and that component contains URL path metacharacters, it breaks out of its intended segment and changes the effective request path.

In 2.97.0, request URLs are constructed in a safe manner, so a variable component can no longer escape its intended segment.

Impact

An attacker who can control a value that flows into a request path can cause gh to send the request to a different API endpoint than intended. This can lead the CLI to read or act on a resource other than the one the user meant to target.

Remediation and Mitigation

1. Upgrade gh to 2.97.0. 2. Be cautious when running gh against untrusted repositories or inputs whose names you do not control.

1 / 3
Source: GitHub
First published (updated )
Severity
5.1
Path Traversal
AV:N/AC:L/AT:N/PR:N/UI:A/VC:N/VI:N/VA:N/SC:L/SI:L/SA:N

GitHub CLI (gh) is GitHub’s official command line tool. Prior to 2.97.0, some HTTP request URLs interpolate variable path components without percent encoding, allowing URL path metacharacters in attacker-controlled repository or resource values to make gh address a different API endpoint or resource than the user intended. This issue is fixed in version 2.97.0.

First published (updated )
Severity
2.1
AV:N/AC:L/AT:P/PR:N/UI:A/VC:N/VI:N/VA:N/SC:N/SI:L/SA:N

GitHub CLI (gh) is GitHub’s official command line tool. Prior to 2.97.0, gh attestation verify  builds the certificate Subject Alternative Name matcher from the --signer-repo and --signer-workflow  flag values without escaping regex metacharacters, so a user-supplied repository or workflow name is treated as a regular expression rather than a literal string. Because GitHub permits characters such as  .  in organization, repository, and workflow path names and  .  is a regex wildcard, an attacker can register a lookalike name (for example github/artifact.attestations-workflows) that satisfies a matcher intended for a different trusted signer (github/artifact-attestations-workflows), bypassing the intended Sigstore attestation verification. Exploitation requires the attacker to create a plausible lookalike repository and produce valid attestations from it, which could undermine supply chain verification for CI/CD pipelines or policy gates that pin trust to a specific signing workflow. This issue is fixed in version 2.97.0.

First published (updated )
Severity
3.5
AV:N/AC:L/PR:L/UI:R/S:U/C:N/I:L/A:N

Summary

A security vulnerability has been identified in GitHub CLI that could allow terminal escape sequence injection when users view GitHub Actions workflow logs using gh run view --log or gh run view --log-failed.

Details

The vulnerability stems from the way GitHub CLI handles raw Actions log output. The gh run view --log and gh run view --log-failed commands stream workflow log lines to stdout or the configured pager without sanitizing terminal control sequences. An attacker who can influence GitHub Actions log content, for example via a PR triggered workflow, can embed escape sequences that are replayed in the user's terminal when they inspect the run.

Depending on the victim's terminal emulator, injected sequences could change the window title, manipulate on screen content, or in some terminal emulators (such as screen) potentially execute arbitrary commands.

In 2.92.0, GitHub CLI sanitizes terminal control sequences in Actions log output before writing to the terminal.

PoC

Create a workflow that emits terminal escape sequences in its log output:

yaml name: Escape Sequence PoC

on: workflowdispatch:

jobs: emit-escape-sequences: runs-on: ubuntu-latest steps: - name: Emit terminal escape sequences run: | # OSC title set printf 'ESCAPEMARKERSTART \033]0;HIJACKEDTITLE\007 ESCAPEMARKEREND\n' # CSI color printf 'ESCAPEMARKERSTART \033[31mREDTEXT\033[0m ESCAPEMARKEREND\n' # Screen title set (enables command execution in screen terminal) printf 'ESCAPEMARKERSTART \033k;malicious command;\033\\ ESCAPEMARKEREND\n'

Then trigger the workflow and view its logs:

bash gh workflow run 'Escape Sequence PoC' gh run view <runid> --log

On vulnerable versions, the raw ESC bytes (0x1b) are passed through to the terminal unsanitized. On 2.92.0 and later, escape sequences are stripped and only the safe visible text is displayed.

Impact

An attacker who can control GitHub Actions workflow output can inject terminal escape sequences into a maintainer's terminal session when they inspect the run with gh run view --log or gh run view --log-failed. The practical impact depends on the victim's terminal emulator.

Remediation and Mitigation

1. Upgrade gh to 2.92.0 2. Pipe log output through a sanitizer (e.g., gh run view --log | cat -v) as a workaround on older versions 3. Exercise caution when viewing logs from untrusted workflow runs

1 / 3
Source: GitHub
First published (updated )
Severity
3.5
AV:N/AC:L/PR:L/UI:R/S:U/C:N/I:L/A:N

gh is GitHub’s official command line tool. From 1.6.0 to before 2.92.0, a security vulnerability has been identified in GitHub CLI that could allow terminal escape sequence injection when users view GitHub Actions workflow logs using gh run view --log or gh run view --log-failed. The vulnerability stems from the way GitHub CLI handles raw Actions log output. The gh run view --log and gh run view --log-failed commands stream workflow log lines to stdout or the configured pager without sanitizing terminal control sequences. An attacker who can influence GitHub Actions log content, for example via a PR triggered workflow, can embed escape sequences that are replayed in the user's terminal when they inspect the run. Depending on the victim's terminal emulator, injected sequences could change the window title, manipulate on screen content, or in some terminal emulators (such as screen) potentially execute arbitrary commands. This vulnerability is fixed in 2.92.0.

First published (updated )
Severity
7.7
Use After Free
AV:N/AC:H/AT:N/PR:L/UI:N/VC:H/VI:H/VA:H/SC:L/SI:N/SA:N/E:X/CR:X/IR:X/AR:X/MAV:X/MAC:X/MAT:X/MPR:X/MUI:X/MVC:X/MVI:X/MVA:X/MSC:X/MSI:X/MSA:X/S:X/AU:X/R:X/V:X/RE:X/U:X

A remote code execution vulnerability was identified in GitHub Enterprise Server in the Git service (babeld). babeld held references into internal collections while calling into the SSH library, where a re-entrant callback could modify a collection and invalidate those references. An attacker with an account and SSH access to the instance could leverage this use-after-free into a controlled write and run arbitrary commands as babeld, including disclosure of its environment secrets. Exploitation required per-build memory offsets; failed attempts crashed babeld. This affected all releases prior to 3.23 and was fixed in 3.22.3, 3.21.8, 3.20.10, and 3.19.14. This vulnerability was reported via the GitHub Bug Bounty program.

First published (updated )
Severity
5.3
AV:N/AC:L/AT:N/PR:L/UI:N/VC:L/VI:N/VA:N/SC:N/SI:N/SA:N/E:X/CR:X/IR:X/AR:X/MAV:X/MAC:X/MAT:X/MPR:X/MUI:X/MVC:X/MVI:X/MVA:X/MSC:X/MSI:X/MSA:X/S:X/AU:X/R:X/V:X/RE:X/U:X

An out-of-bounds read vulnerability was identified in GitHub Enterprise Server in the Git service (babeld). When proving possession of a host key, babeld's SSH implementation sized the session identifier by the current digest length rather than its own. An attacker with an account and SSH access could re-key to an algorithm with a longer hash, causing adjacent process memory to be signed and returned to the client, where a limited amount of it may be recoverable. This vulnerability affected all releases prior to 3.23 and was fixed in versions 3.22.3, 3.21.8, 3.20.10, and 3.19.14. This vulnerability was reported via the GitHub Bug Bounty program.

First published (updated )
Severity
5.3
CSRF
AV:N/AC:L/AT:N/PR:N/UI:P/VC:N/VI:L/VA:L/SC:N/SI:N/SA:N/E:X/CR:X/IR:X/AR:X/MAV:X/MAC:X/MAT:X/MPR:X/MUI:X/MVC:X/MVI:X/MVA:X/MSC:X/MSI:X/MSA:X/S:X/AU:X/R:X/V:X/RE:X/U:X

A cross-site request forgery vulnerability was identified in GitHub Enterprise Server that allowed an unauthenticated attacker to delete stored Slack and Microsoft Teams ChatOps configuration and provider credentials by inducing an authenticated Management Console operator's browser to send a cross-site GET request to the integration removal endpoint without a CSRF token. Exploitation required configured ChatOps integrations and an operator visiting attacker-controlled content while signed in to the Management Console; deletion could disrupt ChatOps notifications and commands when configuration was applied or services reloaded. This vulnerability affected all versions of GitHub Enterprise Server prior to 3.23 and was fixed in versions 3.19.14, 3.20.10, 3.21.8, and 3.22.3. This vulnerability was reported via the GitHub Bug Bounty program.

First published (updated )
Severity
8.8
AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H/E:U/RL:O/RC:C

GitHub Copilot and Visual Studio Code Security Feature Bypass Vulnerability

1 / 2
Source: Microsoft
First published (updated )
Severity
7.8
OS Command Injection, Command Injection
AV:L/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H/E:U/RL:O/RC:C

GitHub Copilot and Visual Studio Code Elevation of Privilege Vulnerability

1 / 2
Source: Microsoft
First published (updated )
Severity
9.8
EPSS
0.42%
Integer Overflow
AV:N/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H

Impact

CommonMarker uses cmark-gfm for rendering Github Flavored Markdown. An integer overflow in cmark-gfm's table row parsing may lead to heap memory corruption when parsing tables who's marker rows contain more than UINT16MAX columns. The impact of this heap corruption ranges from Information Leak to Arbitrary Code Execution.

If affected versions of CommonMarker are used for rendering remote user controlled markdown, this vulnerability may lead to Remote Code Execution (RCE).

Patches

This vulnerability has been patched in the following CommonMarker release:

- v0.23.4

Workarounds

The vulnerability exists in the table markdown extensions of cmark-gfm. Disabling any use of the table extension will prevent this vulnerability from being triggered.

References

- https://github.com/github/cmark-gfm/security/advisories/GHSA-mc3g-88wq-6f4x

Acknowledgements

We would like to thank Felix Wilhelm of Google's Project Zero for reporting this vulnerability

For more information

If you have any questions or comments about this advisory:

Open an issue in CommonMarker

1 / 3
First published (updated )
Severity
6.5
SQL Injection
AV:N/AC:L/PR:N/UI:N/S:U/C:L/I:L/A:N

Github Restaurant Website Restoran v1.0 was discovered to contain a SQL injection vulnerability via the Contact Form page.

First published (updated )
Severity
7.1
EPSS
0.04%
AV:N/AC:L/AT:N/PR:L/UI:N/VC:H/VI:N/VA:N/SC:N/SI:N/SA:N/E:X/CR:X/IR:X/AR:X/MAV:X/MAC:X/MAT:X/MPR:X/MUI:X/MVC:X/MVI:X/MVA:X/MSC:X/MSI:X/MSA:X/S:X/AU:X/R:X/V:X/RE:X/U:X

Impact summary

In some circumstances, debug artifacts uploaded by the CodeQL Action after a failed code scanning workflow run may contain the environment variables from the workflow run, including any secrets that were exposed as environment variables to the workflow. Users with read access to the repository would be able to access this artifact, containing any secrets from the environment.

For some affected workflow runs, the exposed environment variables in the debug artifacts included a valid GITHUBTOKEN for the workflow run, which has access to the repository in which the workflow ran, and all the permissions specified in the workflow or job. The GITHUBTOKEN is valid until the job completes or 24 hours has elapsed, whichever comes first.

Environment variables are exposed only from workflow runs that satisfy all of the following conditions: - Code scanning workflow configured to scan the Java/Kotlin languages. - Running in a repository containing Kotlin source code. - Running with debug artifacts enabled. - Using CodeQL Action versions <= 3.28.2, and CodeQL CLI versions >= 2.9.2 (May 2022) and <= 2.20.2. - The workflow run fails before the CodeQL database is finalized within the github/codeql-action/analyze step. - Running in any GitHub environment: GitHub.com, GitHub Enterprise Cloud, and GitHub Enterprise Server. (Note: artifacts are only accessible to users within the same GitHub environment with access to the scanned repo.)

The GITHUBTOKEN exposed in this way would only have been valid for workflow runs that satisfy all of the following conditions, in addition to the conditions above: - Using CodeQL Action versions >= 3.26.11 (October 2024) and <= 3.28.2, or >= 2.26.11 and < 3. - Running in GitHub.com or GitHub Enterprise Cloud only (not valid on GitHub Enterprise Server).

In rare cases during advanced setup, logging of environment variables may also occur during database creation of Java, Swift, and C/C++. Please read the corresponding CodeQL CLI advisory GHSA-gqh3-9prg-j95m for more details.

Impact details

In CodeQL CLI versions >= 2.9.2 and <= 2.20.2, the CodeQL Kotlin extractor logs all environment variables by default into an intermediate file during the process of creating a CodeQL database for Kotlin code. This is a part of the CodeQL CLI and is invoked by the CodeQL Action for analyzing Kotlin repositories. On Actions, the environment variables logged include GITHUBTOKEN, which grants permissions to the repository being scanned.

The intermediate file containing environment variables is deleted when finalizing the database, so it is not included in a successfully created database. It is, however, included in the debug artifact that is uploaded on a failed analysis run if the CodeQL Action was invoked in debug mode.

Therefore, under these specific circumstances (incomplete database creation using the CodeQL Action in debug mode) an attacker with access to the debug artifact would gain unauthorized access to repository secrets from the environment, including both the GITHUBTOKEN and any user-configured secrets made available via environment variables.

The impact of the GITHUBTOKEN leaked in this environment is limited: - For workflows on GitHub.com and GitHub Enterprise Cloud using CodeQL Action versions >= 3.26.11 and <= 3.28.2, or >= 2.26.11 and < 3, which in turn use the actions/artifacts v4 library, the debug artifact is uploaded before the workflow job completes. During this time the GITHUBTOKEN is still valid, providing an opportunity for attackers to gain access to the repository. - For all other workflows, the debug artifact is uploaded after the workflow job completes, at which point the leaked GITHUBTOKEN has been revoked and cannot be used to access the repository.

Mitigations

Update to CodeQL Action version 3.28.3 or later, or CodeQL CLI version 2.20.3 or later.

Patches

This vulnerability has been fixed in CodeQL Action version 3.28.3, which no longer uploads database artifacts in debug mode. This vulnerability will be fixed in CodeQL CLI version 2.20.3, in which database creation for all languages no longer logs the complete environment by default.

References

- Pull request that bundled CodeQL CLI 2.9.2 with Kotlin extractor environment variable logging - Pull request that introduced the actions/artifacts v4 library, allowing for GITHUBTOKEN exposure in the CodeQL Action debug artifacts before the token was revoked - Related security advisory for the CodeQL CLI

1 / 2
Source: GitHub
First published (updated )
Severity
7.5
EPSS
0.10%
OS Command Injection
AV:N/AC:L/AT:P/PR:N/UI:A/VC:H/VI:H/VA:H/SC:N/SI:N/SA:N/E:X/CR:X/IR:X/AR:X/MAV:X/MAC:X/MAT:X/MPR:X/MUI:X/MVC:X/MVI:X/MVA:X/MSC:X/MSI:X/MSA:X/S:X/AU:X/R:X/V:X/RE:X/U:X

Summary

A security vulnerability has been identified in GitHub Copilot CLI's shell tool that could allow arbitrary code execution through crafted bash parameter expansion patterns. An attacker who can influence the commands executed by the agent (e.g., via prompt injection through repository files, MCP server responses, or user instructions) can exploit bash parameter transformation operators to execute hidden commands, bypassing the safety assessment that classifies commands as "read-only."

Details

The vulnerability stems from how the CLI's shell safety assessment evaluates commands before execution. The safety layer parses and classifies shell commands as either read-only (safe) or write-capable (requires user approval). However, several bash parameter expansion features can embed executable code within arguments to otherwise read-only commands, causing them to appear safe while actually performing arbitrary operations.

The specific dangerous patterns are:

1. ${var@P} — Prompt expansion: The @P parameter transformation operator evaluates its value as a prompt string, which interprets embedded command substitutions. This allows hidden command execution inside what appears to be a simple variable reference.

2. ${var=value} / ${var:=value} — Assignment side-effects: These forms assign values to variables as a side-effect of expansion. When chained with @P, an attacker can progressively build up a command substitution string across multiple expansions.

3. ${!var} — Indirect expansion: Dereferences an arbitrary variable name, which can be combined with other patterns to construct and execute commands dynamically.

4. Nested $(cmd) or <(cmd) inside ${...} expansions: Command substitution or process substitution embedded within parameter expansion default values (e.g., ${HOME:-$(whoami)}) executes the nested command.

Proof of Concept

The following command appears to run a harmless echo, but actually executes touch /tmp/pwned through chained parameter expansion:

bash echo ${a="$"}${b="$a(touch /tmp/pwned)"}${b@P}

How it works: - ${a="$"} assigns the literal $ character to variable a - ${b="$a(touch /tmp/pwned)"} expands $a to $, constructing the string $(touch /tmp/pwned) and assigning it to b - ${b@P} applies prompt expansion to b, which evaluates the embedded $(touch /tmp/pwned) command substitution

Prior to the fix, the safety assessment would classify echo as a read-only command and allow execution without user confirmation — even in modes that normally require approval for write operations.

Impact

An attacker who can influence command text sent to the shell tool — for example, through: - Prompt injection via malicious repository content (README files, code comments, issue bodies) - Compromised or malicious MCP server responses - Crafted user instructions containing obfuscated commands

— could achieve arbitrary code execution on the user's workstation. This is possible even in permission modes that require user approval for write operations, since the commands can appear to be using only read-only utilities to ultimately trigger write operations.

Successful exploitation could lead to data exfiltration, file modification, or further system compromise.

Affected Versions

- GitHub Copilot CLI versions prior to 0.0.423

Remediation and Mitigation

Fix

The fix adds two layers of defense:

1. Parse-time detection: The shell safety assessment analyzes ${...} expansion nodes within bash commands, detecting dangerous operators (@P, =, :=, !) and nested command/process substitutions. Commands containing these patterns are downgraded from read-only to write-capable, ensuring they require user approval.

2. Unconditional blocking: Commands with dangerous expansion patterns are unconditionally blocked at the tool execution layer — regardless of permission mode (including --yolo / autopilot). This prevents exploitation even when all commands are auto-approved.

3. System prompt hardening: The bash shell tool's system prompt now includes explicit instructions for the LLM to refuse executing commands with these patterns, providing a defense-in-depth layer.

User Actions

1. Upgrade GitHub Copilot CLI to 0.0.423 or later. 2. Exercise caution when working in untrusted repositories or with untrusted MCP servers. 3. Review any shell commands suggested by the agent that contain complex parameter expansion patterns.

1 / 2
Source: GitHub
First published (updated )
Severity
9.5
EPSS
0.04%
AV:N/AC:H/AT:P/PR:N/UI:N/VC:H/VI:H/VA:L/SC:H/SI:H/SA:L/E:X/CR:X/IR:X/AR:X/MAV:X/MAC:X/MAT:X/MPR:X/MUI:X/MVC:X/MVI:X/MVA:X/MSC:X/MSI:X/MSA:X/S:X/AU:X/R:U/V:C/RE:H/U:Red

An XML signature wrapping vulnerability was present in GitHub Enterprise Server (GHES) when using SAML authentication with specific identity providers utilizing publicly exposed signed federation metadata XML. This vulnerability allowed an attacker with direct network access to GitHub Enterprise Server to forge a SAML response to provision and/or gain access to a user with site administrator privileges. Exploitation of this vulnerability would allow unauthorized access to the instance without requiring prior authentication. This vulnerability affected all versions of GitHub Enterprise Server prior to 3.14 and was fixed in versions 3.13.3, 3.12.8, 3.11.14, and 3.10.16. This vulnerability was reported via the GitHub Bug Bounty program.

First published (updated )
Severity
8.6
Path Traversal
AV:N/AC:L/AT:N/PR:L/UI:N/VC:H/VI:H/VA:N/SC:N/SI:N/SA:N/E:X/CR:X/IR:X/AR:X/MAV:X/MAC:X/MAT:X/MPR:X/MUI:X/MVC:X/MVI:X/MVA:X/MSC:X/MSI:X/MSA:X/S:X/AU:X/R:X/V:X/RE:X/U:X

A path traversal vulnerability was identified in GitHub Enterprise Server that allowed an attacker with code execution inside a Dependabot updater container to write files to arbitrary repository paths, including GitHub Actions workflow files under .github/workflows/, because path validation did not check the actual write destination. The initial fix in versions 3.21.3, 3.20.5, 3.19.9, 3.18.12, and 3.17.18 was incomplete because validation checked the supplied symlink target, while the publisher wrote the file using its directory and name. Writing GitHub Actions workflow files required the Dependabot installation to have workflows: write permission. Depending on repository permissions and workflow configuration, including a relevant pullrequesttarget workflow or auto-merge, an injected workflow could execute with access to GitHub Actions secrets. This vulnerability affected all versions of GitHub Enterprise Server prior to 3.23 and was fully fixed in versions 3.22.3, 3.21.8, 3.20.10, and 3.19.14. This vulnerability was reported via the GitHub Bug Bounty program.

First published (updated )
Severity
7.5
AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H

cmark-gfm is GitHub's fork of cmark, a CommonMark parsing and rendering library and program in C. Versions prior to 0.29.0.gfm.7 contain a polynomial time complexity issue in handleclosebracket that may lead to unbounded resource exhaustion and subsequent denial of service. This vulnerability has been patched in 0.29.0.gfm.7.

1 / 3
Source: MITRE
First published (updated )

Contact

SecAlerts Pty Ltd.
132 Wickham Terrace
Fortitude Valley,
QLD 4006, Australia
info@secalerts.co
By using SecAlerts services, you agree to our services end-user license agreement. This website is safeguarded by reCAPTCHA and governed by the Google Privacy Policy and Terms of Service. All names, logos, and brands of products are owned by their respective owners, and any usage of these names, logos, and brands for identification purposes only does not imply endorsement. If you possess any content that requires removal, please get in touch with us.
© 2026 SecAlerts Pty Ltd.
ABN: 70 645 966 203, ACN: 645 966 203