Summary
Aerie/PlanDev sequencing-server trusts a caller-provided Hasura session role instead of verifying the role from an authenticated JWT or trusted Hasura request. This can allow unauthenticated attackers to modify expansion rules and command dictionaries, potentially influencing how spacecraft activities are translated into commands.
Affected context
The vulnerability occurs in sequencing-server authorization middleware when request-body session variables are accepted as an authority signal without confirming that the request originated from Hasura. It affects endpoints that use this role check, including expansion-rule insertion, as well as endpoints explicitly exempted from authentication through the endpoint whitelist.
Exploit preconditions
- Attack surface: HTTP endpoints exposed by the sequencing-server
- Required access: No authentication may be required when the affected authorization path is reached
- Trigger point: Authorization processing for expansion-rule or command-dictionary write requests
- Why exploitation is possible: Client-controlled session metadata can satisfy an administrative role check, while a whitelisted write endpoint may bypass authentication entirely
Why this happens
The authorization middleware derives authorization from session variables supplied in the request body rather than exclusively from verified authentication claims or a trusted identity-propagation channel. This is a broken authentication and authorization pattern mapped to CWE-306 because sensitive write operations can be reached without effective authentication; treating unverified client input as a Hasura role also creates an authority confusion between caller-controlled data and trusted identity context.
Potential impact
- An unauthenticated user may insert or alter expansion rules
- Spacecraft activity-to-command translation behavior could be manipulated
- Command dictionaries may be modified without authorization
- Operational workflows or mission planning data could be disrupted
- Malicious rules or dictionary entries could contribute to unsafe command generation
General mitigation guidance
Mitigation generally focuses on making authentication and authorization
decisions exclusively from verified, trusted identity context.
Common defensive patterns include:
- Deriving roles only from validated JWT claims or authenticated server-to-server context
- Rejecting client-supplied session variables as authorization credentials
- Applying explicit authentication and authorization checks to every state-changing endpoint
- Removing unauthenticated write operations from endpoint allowlists
- Enforcing least privilege for roles that can modify sequencing data
- Validating expansion rules and command dictionaries against strict schemas and approved change policies
- Using audit logging, approval workflows, and immutable controls for sensitive operational configuration
How teams detect issues like this
Organizations can look for:
- Static-analysis findings where request-body fields influence role or privilege decisions
- Code-review signals involving authentication bypass allowlists or unauthenticated state-changing routes
- Tests that invoke administrative endpoints without credentials and verify rejection
- Policy checks requiring authentication middleware on all mutation endpoints
- Runtime logs showing anonymous writes, unexpected role claims, or changes to expansion rules and command dictionaries
- Automated security review tools that trace user-controlled data into authorization decisions

