GHSA-55f6-4pr5-c7m5: Kahi has privilege-drop and socket/log permission issues
Kahi releases up to and including v0.1.0-alpha.8 contain three privilege/permission issues, all fixed in v0.1.0-alpha.9. They were identified in a full-codebase security review on 2026-05-26. Affected versions All releases <= v0.1.0-alpha.8. Patched version v0.1.0-alpha.9. Details 1.
Published Jun 30, 2026
Affected packages
github.com/kahiteam/kahi
- Ecosystem
- Go
- Affected
- <= 0.1.0-alpha.8
- Fixed in
- 0.1.0-alpha.9
- Check
go list -m github.com/kahiteam/kahi
Advisory details
From the GitHub Advisory Database (CC BY 4.0).
Kahi releases up to and including v0.1.0-alpha.8 contain three privilege/permission issues, all fixed in v0.1.0-alpha.9. They were identified in a full-codebase security review on 2026-05-26.
Affected versions
All releases <= v0.1.0-alpha.8.
Patched version
v0.1.0-alpha.9.
Details
1. Per-process privilege drop not applied (High)
A process configured with user = "uid:gid" had its credential resolved but never attached to the spawned child, so the process ran with the supervisor's inherited privileges (root, when the supervisor runs as root) instead of the configured lower-privilege user. The intended privilege isolation did not occur, and no error was raised.
2. Privilege drop did not reset supplementary groups (Medium)
When the daemon dropped privileges it set the primary gid and uid but never called setgroups(2), so the launching user's supplementary groups (for example docker, which is root-equivalent) remained active after the drop and were inherited by child processes.
3. FastCGI unix socket world-accessible by default (Medium)
A FastCGI unix-domain socket was chmod-ed only when socket_mode was explicitly configured. With socket_mode unset the socket kept the umask-dependent default (commonly world-accessible), allowing any local user to connect to it.
Remediation
Upgrade to v0.1.0-alpha.9. Privilege handling is now fail-closed: the configured credential is applied or the process refuses to start, supplementary groups are reset via setgroups before setgid/setuid, and FastCGI unix sockets default to 0700.
Workarounds (for <= v0.1.0-alpha.8)
- Do not rely on per-process
user; run the supervisor directly as the intended unprivileged user. - Set an explicit restrictive
socket_modeon FastCGI programs. - Avoid running the supervisor as root where possible.
References
How this page is made: the data panel comes from the GitHub Advisory Database, CISA’s KEV catalog, the NVD and FIRST EPSS, and updates twice a day. No analysis has been written for this entry yet. Methodology
