OS Command Injection in codecov - CVE-2020-15123
Published: July 20, 2020 / Updated: August 8, 2020
codecov
Codecov
Description
The vulnerability allows a remote non-authenticated attacker to read and manipulate data.
In codecov (npm package) before version 3.7.1 the upload method has a command injection vulnerability. Clients of the codecov-node library are unlikely to be aware of this, so they might unwittingly write code that contains a vulnerability. A similar CVE (CVE-2020-7597 for GHSA-5q88-cjfq-g2mh) was issued but the fix was incomplete. It only blocked &, and command injection is still possible using backticks instead to bypass the sanitizer. The attack surface is low in this case. Particularly in the standard use of codecov, where the module is used directly in a build pipeline, not built against as a library in another application that may supply malicious input and perform command injection.
Remediation
External links
- https://github.com/advisories/GHSA-5q88-cjfq-g2mh
- https://github.com/codecov/codecov-node/commit/c0711c656686e902af2cd92d6aecc8074de4d83d
- https://github.com/codecov/codecov-node/pull/180
- https://github.com/codecov/codecov-node/security/advisories/GHSA-xp63-6vf5-xf3v
- https://lgtm.com/query/7714424068617023832