TL;DR: A public Content API oracle, Content-Type Builder query injection, upload-policy bypasses, and refresh-token session retention can combine into account takeover and broader CMS compromise, according to Corgea’s analysis of the Strapi advisory cluster. The practical lesson is that CMS security now depends on hardening exposed request paths, not just patching dependencies.
At a glance
What this is: Corgea says multiple Strapi advisories expose public query oracles, admin-path SQL injection, upload-policy bypasses, and sessions that can survive password changes.
Why it matters: CMS teams need to treat exposed request routes, admin workflows, and session revocation as a single governance problem because one flaw can expose tokens, preserve access, or widen blast radius.
Context
Strapi’s advisory cluster is not just a set of isolated bugs. It shows how a content management system can become dangerous when public content paths, authenticated admin workflows, and session controls intersect in the same application boundary. For IAM and application security teams, the practical issue is whether public routes can reach protected identity data or whether admin actions can be abused through ordinary product features.
The article also shows why CMS hardening is now a governance concern, not just a patching exercise. Public filtering through admin relations, database writes exposed through schema tools, and refresh tokens that survive password changes all affect the control plane around identity and access inside the CMS. That makes the article relevant to teams responsible for authentication, authorization, and privileged workflow protection.
Key questions
Q: What breaks when public CMS filters can reach privileged identity fields?
A: Public filters can become a secret-recovery channel when they traverse into admin-related fields and return different responses for guessed values. That lets an attacker test tokens character by character, which is why query design must be treated as an identity control, not just an application feature.
Q: Why do CMS admin workflows need privileged access controls?
A: Because schema changes, upload settings, and account management can all become attack paths when they are reachable from normal application routes. If an administrator action can influence raw database execution or session state, it deserves the same governance as other high-risk privileged operations. That is a privileged access problem, not just a feature design choice.
Q: How do security teams know if refresh-token revocation is failing?
A: If old sessions still work after a password reset or reset-token rotation, revocation is not operating as a true containment control. Teams should test whether credential changes force a clean session boundary, then confirm that stale refresh tokens cannot authenticate after recovery actions. Without that test, attackers may keep access after the password is changed.
Q: What should CMS teams do when uploads can include active content?
A: Treat active uploads as potential execution payloads, not inert files. Separate upload hosting from the admin origin, restrict risky file types, and verify that an administrator cannot load attacker-controlled HTML or SVG inside a trusted CMS session. That reduces the chance that uploaded content becomes a browser-side pivot into privileged actions.
Technical breakdown
Public Content API filters can become a boolean oracle
A boolean oracle appears when a public endpoint reveals whether a guessed value is right or wrong through a change in response behaviour. In this case, relational filters can traverse from public content records into restricted admin-user fields, so a series of crafted requests can test candidate token values one character at a time. The attacker does not need direct admin access, only a public content type with a relation that crosses trust boundaries. That turns filtering logic into an information-disclosure channel rather than a simple query feature.
Practical implication: review public filter paths for any relation that can reach admin-owned fields and block cross-boundary queries at the authorization layer.
Content-Type Builder writes can reach database execution paths
Schema tools often appear administrative rather than dangerous, but they can become high impact when user-controlled values are passed into raw database primitives. Here, a default value tuple can flow into Knex raw execution during schema migration, which means the application is no longer just storing metadata. Depending on the database engine and deployment posture, that path can support arbitrary SQL statements, file read primitives, denial of service, or even database-server code execution. Restricting the route to development mode removes the write surface from production.
Practical implication: keep schema mutation interfaces out of production and treat any admin feature that can alter database structure as privileged attack surface.
Refresh-token retention and upload-policy bypasses expand the blast radius
Session invalidation and content handling are part of the same containment story. If a password change does not revoke existing refresh-token sessions, an attacker can keep using a live session after credentials are reset. Separately, if upload policy enforcement can be bypassed for active content such as HTML or SVG, an attacker can place files that may execute in an administrator’s browser when opened from the CMS origin. Those two failures make compromise harder to contain because access and payload delivery both survive longer than they should.
Practical implication: enforce token revocation on credential reset and isolate uploaded active content from the admin origin.
Threat narrative
Attacker objective: The attacker aims to recover privileged CMS data, preserve authenticated access, or convert administrative reach into broader control of the Strapi instance.
- Entry begins with a public Content API request that probes relational filters against admin-user fields.
- Credential access follows when repeated boolean queries recover values such as admin reset-password tokens or other sensitive attributes.
- Privilege retention or impact can follow through preserved refresh-token sessions, authenticated admin abuse, or admin-side execution of uploaded active content.
Breaches seen in the wild
- Gladinet Hard-Coded Keys RCE Exploitation: Actively exploited hard-coded keys in Gladinet CentreStack and Triofox enable remote code execution.
- LiteLLM MCP auth bypass 2026: An exploited LiteLLM MCP auth bypass and default sk-1234 master keys let attackers steal AI gateway master and provider API keys.
Read and download The State of NHI & AI Agent Breach Report 2026, covering 200+ breaches impacting Non-Human Identities including AI Agents.
NHI Mgmt Group analysis
Public filtering is now a governance boundary, not a convenience feature. Strapi’s advisory cluster shows that a CMS query layer can leak identity data when relational filters are allowed to cross from public records into admin-owned fields. That is a control failure, not just a bug class. Security teams should treat exposed filter paths as part of the authorization model, because the query interface itself can become the attack surface.
Schema administration belongs in the privileged change plane. The Content-Type Builder issue matters because it shows how a seemingly routine admin workflow can reach raw database execution. Once application-level schema changes are able to influence database statements, the control problem shifts from app security to privileged operational design. Practitioners should regard schema mutation capability as equivalent to high-risk privileged access, not ordinary configuration.
Session revocation is a containment control, not a housekeeping task. If refresh tokens remain valid after password changes, credential reset no longer ends the incident. That creates a gap between identity recovery and access recovery, which is exactly where attackers keep operating. The broader lesson is that authentication and revocation logic must be designed as a single lifecycle, especially for admin users and CMS operators.
Active upload handling creates a secondary execution plane. When HTML or SVG can bypass upload policy, the CMS can be used to stage browser-executed content inside the admin trust zone. That is a combined application and identity problem because the payload only matters once an administrator loads it. The practical conclusion is that upload controls, origin separation, and admin browser exposure have to be governed together.
Request-path hardening is the named concept this advisory cluster exposes. The important issue is not simply that Strapi has multiple CVEs, but that public request paths, admin workflows, and session continuity all need separate governance. Teams that only patch packages miss the deeper pattern: the dangerous condition is reachable request-path abuse across the CMS control plane. Practitioners should review every public and privileged route for unintended reach into identity-bearing objects.
What this signals
Request-path hardening is the real control question here: CMS teams need to decide whether public query paths, admin mutations, and session lifecycle controls are governed as one trust boundary or treated as separate features. Strapi’s advisory cluster shows why that separation fails in practice.
The operational signal is clear. When a CMS can leak identity data through filters, keep sessions alive after password changes, or accept active uploads in the admin trust zone, the issue is no longer simple patch hygiene. Teams should review authorization, revocation, and upload isolation together rather than as isolated tickets.
For practitioners
- Harden public filter routes against admin-field traversal Block relational queries from public Content API endpoints when they can reach admin-user fields such as reset-password tokens, passwords, or email-linked identity attributes.
- Move schema mutation out of production reach Restrict Content-Type Builder write APIs to development mode and internal networks so administrators cannot trigger schema changes from production traffic.
- Revoke sessions on password change and reset Force refresh-token invalidation whenever administrator passwords or reset tokens change, and verify that old sessions cannot survive the reset flow.
- Isolate active uploads from the admin origin Serve uploaded content from a separate origin and block executable file types such as HTML and SVG from paths that administrators can browse in-session.
- Audit for probing and post-reset compromise Review logs for repeated relational-filter probes, unexpected password-reset events, and admin logins that follow token-oracle style requests.
Key takeaways
- Strapi’s advisory cluster shows how public filtering, admin schema tools, and session handling can combine into a broader CMS compromise path.
- The central weakness is not a single bug class but the overlap between request-path access, identity-bearing data, and privileged workflows.
- Teams should treat relational query limits, session revocation, and upload isolation as part of the same CMS control boundary.
Standards & Framework Alignment
This section maps relevant standards and security frameworks to the operational risks and controls described in this guidance.
OWASP API Security Top 10 addresses the attack and risk surface, while NIST CSF 2.0, NIST SP 800-53 Rev 5, CIS Controls v8 and OWASP ASVS set the governance and control requirements practitioners need to meet.
| Framework | Control / Reference | Relevance |
|---|---|---|
| OWASP API Security Top 10 | API8 — Security Misconfiguration | The public filter and route exposure reflects API and application misconfiguration. |
| Recommendation — Review exposed CMS routes against API8 and remove any public path that can reach privileged data or actions. | ||
| NIST CSF 2.0 | PR.AA-05 — Access Permissions, Entitlements and Authorizations | The advisory cluster centers on broken authorization across public and admin CMS paths. |
| Recommendation — Apply PR.AA-05 to stop public requests from traversing into restricted admin objects or workflows. | ||
| NIST SP 800-53 Rev 5 | AC-6 — Least Privilege | Admin functions and token lifecycle actions should not be reachable beyond least privilege. |
| Recommendation — Enforce AC-6 so only explicitly authorized users can trigger schema changes, token resets, or upload exceptions. | ||
| CIS Controls v8 | CIS-5 — Account Management | The article highlights session retention and admin account recovery weaknesses. |
| Recommendation — Use CIS-5 to govern admin account lifecycle, reset workflows, and session invalidation after credential changes. | ||
| OWASP ASVS | V8 — Authorization | The core flaw is authorization failure across filtered public content and privileged admin routes. |
| Recommendation — Apply V8 to verify that authorization is checked before any relational traversal reaches protected objects. | ||
Key terms
- Boolean Oracle: A Boolean oracle is an API or endpoint that reveals useful information through a simple true or false response. Attackers use it to confirm whether an identity exists, whether a resource is valid, or whether a condition has been met, then automate that check across large datasets for profiling or fraud.
- Session revocation: The ability to invalidate active sessions so access ends immediately instead of waiting for tokens or browser state to expire. For identity governance, this is the control that determines whether authentication still matters after a compromise is detected.
- Privileged Change Plane: The privileged change plane is the set of administrative workflows that can alter application behaviour, schema, permissions, or access state. In practice, these paths carry higher risk than ordinary user features and should be isolated, audited, and tightly authorized.
- Active Upload Content: Active upload content is a file that can execute code or script in a browser or similar runtime, such as HTML or SVG. When such content is accepted into a trusted origin, it can become a delivery mechanism for session hijack, privilege abuse, or admin-side action.
Deepen your knowledge
NHI Mgmt Group covers identity security, NHI governance, and agentic AI through independent research, practitioner guides, and the NHI Foundation Level course. It helps security and identity practitioners strengthen governance across credentials, lifecycle controls, and privileged access patterns.
Published by the NHIMG editorial team on August 21, 2026.
Updated on October 12, 2026.
NHI Mgmt Group, the independent authority on Non-Human Identity, IAM, and Agentic AI security. nhimg.org