Summary
An information disclosure flaw in Capgo's Supabase PostgREST global_stats endpoint allows unauthenticated users with only the public API key to access sensitive financial and operational metrics.
By querying the /rest/v1/global_stats endpoint, remote attackers can retrieve MRR, revenue breakdowns, customer counts, and telemetry that should typically be restricted to internal stakeholders.
Affected context
This vulnerability affects Capgo deployments that expose the Supabase PostgREST /rest/v1/global_stats endpoint and rely on a public API key for access control.
It is most relevant when the endpoint is reachable over the internet, is not protected by additional authentication or authorization checks, and returns detailed financial or operational metrics by design.
Environments that treat the public API key as non-secret or embed it in client-side applications are particularly exposed because it can be trivially obtained by external parties.
Exploit preconditions
- Attack surface: Publicly exposed HTTP(S) API endpoint
/rest/v1/global_statsserved via Supabase PostgREST - Required access: Knowledge or capture of the public API key, which may be embedded in front-end code or documentation
- Authentication state: No per-user authentication or role-based authorization enforced for this endpoint
- Trigger point: Sending crafted HTTP requests to the global_stats endpoint using the public API key in headers or query parameters
- Why exploitation is possible: The endpoint is misclassified as publicly readable and returns sensitive business metrics without enforcing appropriate authorization boundaries
Why this happens
The core issue is an access control and data classification failure where an API endpoint that returns sensitive business metrics is treated as publicly accessible.
This maps to CWE-200: Exposure of Sensitive Information to an Unauthorized Actor, where data intended for restricted audiences is served to any caller meeting weak or incorrect access criteria.
The design implicitly assumes that the public API key offers sufficient protection or that the underlying data is non-sensitive, but the actual response payload includes financial indicators (MRR, total revenue, plan-tier breakdowns) and operational telemetry that should be bound to privileged roles.
Lack of fine-grained authorization checks on the PostgREST layer, combined with permissive row-level or RPC policies, leads to overexposure of internal analytics.
Potential impact
- Competitors could analyze MRR trends, total revenue, and plan-tier breakdown to infer business performance and strategy.
- Attackers may profile customer base size and growth to better target social engineering or account takeover campaigns.
- Public knowledge of revenue and operational telemetry could impact negotiation posture with partners, investors, or vendors.
- Disclosure of operational metrics may reveal deployment scale, regional usage patterns, or system behavior useful for planning further attacks.
- Loss of customer and stakeholder trust due to inadvertent exposure of internal business metrics.
General mitigation guidance
Effective mitigation centers on correctly classifying the sensitivity of analytics endpoints and enforcing matching authorization controls.
Common defensive patterns include:
- Treating financial and operational metrics as privileged data, only accessible to authenticated and authorized internal roles.
- Applying role-based access control (RBAC) or row-level security (RLS) policies to analytics and statistics endpoints, ensuring least-privilege access to aggregated metrics.
- Splitting public, non-sensitive statistics (e.g., anonymized or heavily aggregated numbers) from private, detailed business metrics and exposing them via separate, clearly scoped endpoints.
- Avoiding reliance on public API keys as a security boundary; public keys should be considered identifiers, not secrets.
- Implementing a data classification process so that new endpoints returning internal metrics are reviewed for exposure risk before being made public.
How teams detect issues like this
Organizations can detect similar information disclosure issues by:
- Reviewing API specifications and database schemas to identify endpoints that return financial or operational metrics, then verifying their access control policies.
- Using static analysis or configuration scanning to flag PostgREST endpoints or policies that allow unrestricted SELECT access on sensitive tables or views.
- Implementing automated tests that attempt to access analytics endpoints with anonymous or low-privilege tokens and fail builds when sensitive fields are exposed.
- Monitoring access logs for anonymous or unusually frequent requests to analytics and statistics endpoints that originate from untrusted networks.
- Periodically running external security assessments or bug bounty tests focused on misconfigured public endpoints and over-broad API access.

