This CVE reveals a structural flaw in how Semaphore handles git operations: project-scoped RBAC permissions don't prevent infrastructure-level command execution. The --upload-pack injection in git_url processing sits at a shared infrastructure layer rather than within project sandboxes, meaning any user with Manager or Owner role on any project can reach host-level RCE. The permission model is working exactly as designed—it's just designed around assumptions that became false when multi-tenant deployments became the norm. Originally, 'authenticated user' meant a trusted internal developer. Now it means a user with project-level access in a shared SaaS environment, but the git_url execution path never got re-audited against that shifted threat model. The CVSS 8.8 captures severity but obscures the actual exploitability constraints: this isn't a remotely exploitable flaw without authentication, and the population of users who can trigger it (project Managers/Owners) is typically smaller and more vetted than generic 'authenticated users.' That doesn't excuse the flaw—it suggests our severity metrics for CI/CD argument injection may be systematically miscalibrated. The fix isn't just sanitizing git_url input. The real question is whether project-scoped roles should ever have direct paths to host-level execution on shared infrastructure, or whether that access should require explicit infrastructure-admin separation. The --upload-pack injection pattern has been documented across CI/CD systems since at least 2017, yet each vendor patches at their own layer without architectural knowledge transfer. This isn't just accumulated technical debt—it's compounding exposure where each new CVE in this class teaches attackers the pattern. The remediation that matters is treating git_url processing as a security-critical, continuously-audited code path going forward, not just patching this specific injection vector.
CVE-2026-73682
Official description Straight from the sourceThe vendor's or NVD's own wording, published unedited. Authoritative, but often terse — it says what broke, rarely what to do.
NVD · uneditedSemaphore versions prior to 2.18.20 contain an OS command injection (argument injection) vulnerability in the repository git_url handling that allows authenticated users holding the Manager or Owner role on any project to achieve remote code execution on the Semaphore server host. Attackers can craft a malicious git_url value using git's --upload-pack= option to inject and execute arbitrary shell commands when the server processes repository operations using the default cmd_git client.
Technical summary Written by usOur analysis, written from the advisory, the CVSS vector and the affected-version data. It adds context the advisory leaves out, and never invents facts that are not in the source.
dbcve analysisA detailed technical summary for this CVE is being prepared.
CVSS breakdown How the score is builtThe industry scoring standard. It rates how the flaw is reached, what it takes to exploit, and what an attacker gains — the score is derived from those, not the other way round.
From the vector- Attack vector
- Network
- Complexity
- Low
- Privileges
- Low
- User interaction
- None
- Scope
- Unchanged
- Confidentiality
- High
- Integrity
- High
- Availability
- High
CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H
Remediation Closing itWhat it takes to close this. Where a vendor fix exists we point at it; where none exists we say so plainly, and can build one. Effort estimates are scoped from the advisory, not from your codebase.
dbcve · scoped2.18.20
- Upgrade Semaphore to version 2.18.20 or later to remediate the OS command injection vulnerability in repository git_url handling
Generated from the published advisory — verify against the referenced sources before acting.
Scan for this in your stack
Free · runs locallyCheck whether your project pulls in CVE-2026-73682 — or any other known-vulnerable package — straight from your lock files. Free and open source; it runs locally and uploads nothing.
References Go to the primary sourcePrimary sources — vendor advisories, patches and trackers. Where our summary and a reference disagree, the reference wins.
Primary sourcesThis CVE reveals a structural flaw in how Semaphore handles git operations: project-scoped RBAC permissions don't prevent infrastructure-level command execution. The --upload-pack injection in git_url processing sits at a shared infrastructure layer rather than within project sandboxes, meaning any user with Manager or Owner role on any project can reach host-level RCE. The permission model is working exactly as designed—it's just designed around assumptions that became false when multi-tenant deployments became the norm. Originally, 'authenticated user' meant a trusted internal developer. Now it means a user with project-level access in a shared SaaS environment, but the git_url execution path never got re-audited against that shifted threat model. The CVSS 8.8 captures severity but obscures the actual exploitability constraints: this isn't a remotely exploitable flaw without authentication, and the population of users who can trigger it (project Managers/Owners) is typically smaller and more vetted than generic 'authenticated users.' That doesn't excuse the flaw—it suggests our severity metrics for CI/CD argument injection may be systematically miscalibrated. The fix isn't just sanitizing git_url input. The real question is whether project-scoped roles should ever have direct paths to host-level execution on shared infrastructure, or whether that access should require explicit infrastructure-admin separation. The --upload-pack injection pattern has been documented across CI/CD systems since at least 2017, yet each vendor patches at their own layer without architectural knowledge transfer. This isn't just accumulated technical debt—it's compounding exposure where each new CVE in this class teaches attackers the pattern. The remediation that matters is treating git_url processing as a security-critical, continuously-audited code path going forward, not just patching this specific injection vector.
Practitioner notes
ContributedPeer-ranked notes from engineers who’ve handled CVE-2026-73682 in production — separate from our analysis above.
The advisory tells you what broke. It rarely tells you what actually worked. If you’ve dealt with this one, that detail is what the next engineer is searching for.
- The version that genuinely resolved it — not the one the vendor claimed
- A config change or rule that shut the vector down
- A gotcha in the upgrade path that cost you an afternoon
A place for practitioners to share what actually worked: a mitigation you’ve tested, a configuration change, a version- or environment-specific caveat, or a link to a verified patch. The most useful notes rise to the top as peers upvote them, so the signal stays high.
- Verified mitigations, workarounds, and config changes
- Version or environment caveats, and links to real fixes
- No weaponised exploit code, or anything meant to cause harm
- No spam, self-promotion, credentials, or personal data