GHSA-x287-5c68-36wp is a medium-severity (CVSS 5.3) vulnerability in openwisp-ipam. O3 Security confirms whether GHSA-x287-5c68-36wp is actually reachable in your code before you act, and blocks exploitation at runtime until you patch.
OpenWISP IPAM has broken object-level authorization: ExportSubnetView lets a member of one organization export another organization's subnet and all its IP addresses
Real-World Exposure
openwisp-ipamReal-time download stats are indexed for npm and PyPI packages. This vulnerability affects PyPI packages — download data is not available via public APIs for these ecosystems.
Description
Summary
OpenWISP IPAM is multi-tenant: every Subnet belongs to an organization, and API access is scoped to the organizations a user belongs to. The CSV export endpoint, ExportSubnetView, omits the organization-membership check that its import sibling performs, and loads the subnet by primary key with no organization filter. An authenticated user who is a member of one organization can therefore export a subnet belonging to another organization — its name, CIDR, organization slug and every IP address in it — by issuing an export request for that subnet's id.
Details
ImportSubnetView.post() authorizes first, by calling assert_organization_permissions():
# openwisp_ipam/api/views.py (ImportSubnetView)
def post(self, request, *args, **kwargs):
self.assert_organization_permissions(request) # <-- org-membership check
file = request.FILES["csvfile"]
...
self.subnet_model().import_csv(file)
# openwisp_ipam/api/utils.py:16 (AuthorizeCSVImport)
def assert_organization_permissions(self, request):
if request.user.is_superuser:
return
# ... otherwise verifies the user belongs to the target organization
ExportSubnetView.post() performs no such check. It is declared with IpAddressOrgMixin (whose organization scoping lives in get_queryset()), but its overridden post() never calls get_queryset() or assert_organization_permissions() — it goes straight to export_csv():
# openwisp_ipam/api/views.py:246
class ExportSubnetView(ProtectedAPIMixin, IpAddressOrgMixin, CreateAPIView):
subnet_model = Subnet
def post(self, request, *args, **kwargs):
response = HttpResponse(content_type="text/csv")
response["Content-Disposition"] = 'attachment; filename="ip_address.csv"'
writer = csv.writer(response)
self.subnet_model().export_csv(kwargs["subnet_id"], writer) # no org check reached
return response
export_csv() resolves the subnet by primary key with no organization filter and writes its full contents:
# openwisp_ipam/base/models.py:242
def export_csv(self, subnet_id, writer):
ipaddress_model = load_model("openwisp_ipam", "IpAddress")
subnet = load_model("openwisp_ipam", "Subnet").objects.get(pk=subnet_id) # any org's subnet
# ... writes subnet name, CIDR, organization slug, and every IpAddress (address, description, …)
The subnet_id is a random UUIDField (migrations/0001_initial.py: models.UUIDField(default=uuid.uuid4)), so an attacker must obtain the target subnet's id rather than enumerate it — hence Attack Complexity: High. This is a genuine missing-authorization flaw regardless: the import/export asymmetry shows the check was intended, and subnet ids routinely appear in URLs, logs, shared CSV exports and support tickets, from which a cross-organization id can leak.
The same single-object pattern (resolving the subnet by subnet_id outside the org-scoped queryset) also appears in AvailableIpView.get() and RequestIPView.post() (the latter additionally writes an IP into the target subnet); these are worth reviewing alongside the export fix.
Proof of concept
Prerequisites: an OpenWISP IPAM instance with two organizations OrgA and OrgB; a normal (non-superuser) user who is a member of OrgB only; a subnet in OrgA whose id S is known to the attacker (e.g. leaked via a URL/log/shared export).
- As the OrgB user, authenticate to the API.
POST /api/v1/subnet/{S}/export/(OrgA's subnet id).- Result: a CSV download containing OrgA's subnet name, CIDR, organization slug and every IP address in it — even though the user belongs only to OrgB.
- Control (proves it's an oversight): the corresponding import endpoint
POST /api/v1/import-subnet/with OrgA data is rejected for the same user byassert_organization_permissions().
(Reported from a first-hand source review of the current master: the missing check in ExportSubnetView.post, the guarded ImportSubnetView.post sibling, the un-filtered Subnet.objects.get(pk=…) in export_csv, and the UUIDField primary key were each verified by hand. The fix is a one-line authorization call, so a maintainer can confirm trivially with two organizations.)
Impact
A member of one organization can read the complete contents of another organization's subnet — its IP allocation inventory (addresses, descriptions, organization slug, CIDR). In a WISP / network-management deployment this discloses another tenant's network topology and host inventory. Exploitation requires obtaining the target subnet's UUID (AC:H), but the authorization check is missing outright. (RequestIPView shares the pattern and additionally allows writing an IP into another organization's subnet — a separate integrity concern for the maintainer to confirm.)
Remediation
Add the organization-membership check to ExportSubnetView.post() — call self.assert_organization_permissions(request) (as ImportSubnetView does), or resolve the subnet through the org-scoped get_queryset() / get_object() and return 404 when it isn't in the caller's organizations. Apply the same to AvailableIpView and RequestIPView.
Affected Packages
| Ecosystem | Package | Vulnerable range | Fix |
|---|---|---|---|
| 🐍PyPI | openwisp-ipam | all versions | 1.2.1 |
Detection & mitigation playbook
Open-source dependencyDetect
Scan your dependency tree (package-lock.json, pnpm-lock.yaml, requirements.txt, go.sum, etc.) for openwisp-ipam. 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 openwisp-ipam to 1.2.1 or later, then make sure no transitive (indirect) dependency still pins the vulnerable range — O3 confirms GHSA-x287-5c68-36wp 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-x287-5c68-36wp 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-x287-5c68-36wp. 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-x287-5c68-36wp in your dependencies?
O3 detects GHSA-x287-5c68-36wp across PyPI dependencies and uses function-level reachability to confirm whether the vulnerable code path is actually reachable — not just present. No false positives.