Skip to content

Injectable expression expansions in action steps

Low
webknjaz published GHSA-vxmw-7h4f-hqxh Sep 4, 2025

Package

actions pypa/gh-action-pypi-publish (GitHub Actions)

Affected versions

<= v1.12.4

Patched versions

> v1.12.4

Description

Summary

gh-action-pypi-publish makes use of GitHub Actions expression expansions (i.e. ${{ ... }}) in contexts that are potentially attacker controllable. Depending on the trigger used to invoke gh-action-pypi-publish, this may allow an attacker to execute arbitrary code within the context of a workflow step that invokes gh-action-pypi-publish.

Details

gh-action-pypi-publish contains a composite action step, set-repo-and-ref, that makes use of expression expansions:

  - name: Set repo and ref from which to run Docker container action
    id: set-repo-and-ref
    run: |
      # Set repo and ref from which to run Docker container action
      # to handle cases in which `github.action_` context is not set
      # https://github.com/actions/runner/issues/2473
      REF=${{ env.ACTION_REF || env.PR_REF || github.ref_name }}
      REPO=${{ env.ACTION_REPO || env.PR_REPO || github.repository }}
      REPO_ID=${{ env.PR_REPO_ID || github.repository_id }}
      echo "ref=$REF" >>"$GITHUB_OUTPUT"
      echo "repo=$REPO" >>"$GITHUB_OUTPUT"
      echo "repo-id=$REPO_ID" >>"$GITHUB_OUTPUT"
    shell: bash
    env:
      ACTION_REF: ${{ github.action_ref }}
      ACTION_REPO: ${{ github.action_repository }}
      PR_REF: ${{ github.event.pull_request.head.ref }}
      PR_REPO: ${{ github.event.pull_request.head.repo.full_name }}
      PR_REPO_ID: ${{ github.event.pull_request.base.repo.id }}

Permalink:

- name: Set repo and ref from which to run Docker container action
id: set-repo-and-ref
run: |
# Set repo and ref from which to run Docker container action
# to handle cases in which `github.action_` context is not set
# https://github.com/actions/runner/issues/2473
REF=${{ env.ACTION_REF || env.PR_REF || github.ref_name }}
REPO=${{ env.ACTION_REPO || env.PR_REPO || github.repository }}
REPO_ID=${{ env.PR_REPO_ID || github.repository_id }}
echo "ref=$REF" >>"$GITHUB_OUTPUT"
echo "repo=$REPO" >>"$GITHUB_OUTPUT"
echo "repo-id=$REPO_ID" >>"$GITHUB_OUTPUT"

In normal intended operation, these expansions are used to establish a correct priority for outputs like ref and repo-id.

However, these expansions have a side effect: because they're done with ${{ ... }} and not with ${...} (i.e. normal shell interpolation), they can bypass normal shell quoting rules. In particular, if both env.ACTION_REF and env.PR_REF evaluate to empty strings, then the expression falls back to github.ref_name, which can be an attacker controlled string via a branch or tag name.

For example, if the attacker is able to set a branch name to something like innocent;cat${IFS}/etc/passwd, then the REF line may expand as:

REF=innocent;cat${IFS}/etc/passwd

which would set REF to innocent and then run the attacker's code.

Additional information about dangerous expansions can be found in zizmor's template-injection rule documentation.

Impact

The impact of this vulnerability is very low: the expression in question is unlikely to be evaluated in normal operation, since env.ACTION_REF should always take precedence.

In particular, the action is not vulnerable in many popular configurations, i.e. those where pull_request or release or a push: tags event is used to call the action.

Severity

Low

CVSS overall score

This score calculates overall vulnerability severity from 0 to 10 and is based on the Common Vulnerability Scoring System (CVSS).
/ 10

CVSS v3 base metrics

Attack vector
Network
Attack complexity
Low
Privileges required
None
User interaction
None
Scope
Unchanged
Confidentiality
None
Integrity
None
Availability
None

CVSS v3 base metrics

Attack vector: More severe the more the remote (logically and physically) an attacker can be in order to exploit the vulnerability.
Attack complexity: More severe for the least complex attacks.
Privileges required: More severe if no privileges are required.
User interaction: More severe when no user interaction is required.
Scope: More severe when a scope change occurs, e.g. one vulnerable component impacts resources in components beyond its security scope.
Confidentiality: More severe when loss of data confidentiality is highest, measuring the level of data access available to an unauthorized user.
Integrity: More severe when loss of data integrity is the highest, measuring the consequence of data modification possible by an unauthorized user.
Availability: More severe when the loss of impacted component availability is highest.
CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:N

CVE ID

No known CVE

Weaknesses

Improper Neutralization of Special Elements used in a Command ('Command Injection')

The product constructs all or part of a command using externally-influenced input from an upstream component, but it does not neutralize or incorrectly neutralizes special elements that could modify the intended command when it is sent to a downstream component. Learn more on MITRE.

Credits