CVE-2026-47706: Strawberry GraphQL QueryDepthLimiter Denial of Service

Technical summary

The vulnerability exists in the QueryDepthLimiter extension's determine_depth function, which lacks proper cycle detection when evaluating fragment spreads. When a query contains circular references between fragments, the recursive depth calculation never terminates, triggering a RecursionError that halts the validation pipeline. This is a classic algorithmic flaw where defensive programming—specifically, tracking visited nodes to prevent re-entry—was omitted from the traversal logic.

Business impact

Any production GraphQL service running Strawberry GraphQL in the vulnerable version range can be trivially taken offline by an unauthenticated attacker sending a single malicious query. Since no authentication is required and the attack is network-accessible, the business impact is service unavailability without resource-intensive DDoS infrastructure. For API-dependent applications, this translates to customer-facing downtime and potential SLA violations.

Affected systems

Strawberry GraphQL versions 0.71.0 through 0.315.6 are affected. Any GraphQL API or service built on these versions using the QueryDepthLimiter extension is vulnerable. This includes both direct library usage and any framework or SDK that depends on Strawberry GraphQL in the affected range. Verify your dependency chain for transitive Strawberry GraphQL inclusions.

Exploitability

Exploitability is straightforward: no authentication, privileges, or user interaction are required. An attacker only needs network access to the GraphQL endpoint and the ability to submit a query with circular fragment references. CVSS vector AV:N/AC:L/PR:N/UI:N indicates low complexity and high accessibility. Practical exploit code is trivial to construct and requires no special tooling beyond standard GraphQL clients.

Remediation

Upgrade Strawberry GraphQL to version 0.315.7 or later. This version includes cycle detection in the QueryDepthLimiter extension's determine_depth function, preventing infinite recursion on circular fragments. Review your dependency management to identify all services using Strawberry GraphQL and prioritize patching based on traffic and criticality.

Patch guidance

Update to Strawberry GraphQL 0.315.7 or a later stable release. This is a straightforward dependency upgrade; no configuration changes are required. Test the upgraded version in a staging environment to verify compatibility with your GraphQL schema and extensions. Monitor application logs after deployment to confirm validation behavior remains as expected. If you are unable to upgrade immediately, implement network-level query complexity limits or WAF rules to reject deeply nested or circular fragment queries as a temporary compensating control.

Detection guidance

Monitor for RecursionError or RuntimeError exceptions in your GraphQL validation logs. Enable detailed logging in the Strawberry GraphQL validation layer to capture query structures that trigger errors. Analyze incoming queries for circular fragment references or unusual nesting patterns using query analysis tools. Network IDS/IPS signatures targeting GraphQL recursive patterns may help detect attack attempts, though signature-based detection is less effective than dependency tracking and version monitoring.

Why prioritize this

Although the CVSS score is MEDIUM (5.3), prioritization should be HIGH for internet-facing or production GraphQL APIs because exploitability is trivial and impact is immediate service disruption. Organizations running Strawberry GraphQL should patch within 1–2 weeks depending on deployment complexity and testing requirements. Internal or non-critical GraphQL services can follow standard patch cycles.

Risk score, explained

The CVSS score of 5.3 reflects availability impact (A:L) with no confidentiality or integrity effects, low attack complexity, and no privilege or user interaction barriers. The 'L' in availability acknowledges that while the denial of service is real and network-accessible, it is localized to the validation layer and does not necessarily compromise underlying data or systems permanently. However, operational risk is high due to ease of exploitation.

Frequently asked questions