IT SBIT Services
Schneeberger
DEEN
CVE tracker

pgEdge RAG Server: Resource exhaustion via unbounded corpus read

CVE-2026-125327High 7.5

← All vulnerabilities

Summary

pgEdge RAG Server before 2.0.0 reads the entire filtered document corpus from PostgreSQL without a LIMIT and rebuilds an in-memory BM25 keyword index on every query. Hybrid search is enabled by default and the query endpoint has neither authentication nor rate limiting. An unauthenticated remote attacker can send cheap requests that cause heavy database and CPU load and can exhaust service resources, leading to an outage; confidentiality and integrity are not affected.

Affected

  • pgEdge RAG Server (pgedge-rag-server) versions before 2.0.0

Fixed in

  • Upgrade to pgEdge RAG Server 2.0.0 or later

Workaround

  • Unknown: no vendor-documented workaround. Mitigation options: restrict the query endpoint to internal or VPN clients and add rate limiting at a reverse proxy or API gateway.

Exploitation

  • Remote, unauthenticated, low complexity, no user interaction required.
  • Not listed in CISA KEV; OffSeq lists it as not known to be exploited.
  • No public proof of concept from reputable sources found.
  • No reports of in-the-wild attacks found.

Our recommendation

  • Upgrade to pgEdge RAG Server 2.0.0 or later.
  • Until the upgrade is done, do not expose the query endpoint to the internet and add rate limiting.
  • Check whether any client exposes the endpoint and monitor query volume and database and CPU load.
  • Priority is medium to high if the endpoint is exposed, and low if it is internal only.

Public exploits

PoC-in-GitHub, Metasploit and Exploit-DB list nothing for this CVE.

Research sources

Researched and written up in English and German by an AI (Claude by Anthropic) based on NVD, CISA, EPSS and the sources listed. The vendor information is authoritative. Researched on 11 Nov 2026, 05:14.

Original description (NVD)

pgEdge RAG Server is an API server for retrieval-augmented generation over PostgreSQL content using pgvector. Prior to 2.0.0, the unauthenticated query endpoint uses FetchDocuments in internal/database/search.go to read the entire filtered corpus without a LIMIT and rebuild an in-memory BM25 index on every request. Because HybridEnabled is set by default in internal/config/loader.go and the endpoint has no rate limit, inexpensive remote queries can impose database and CPU work proportional to corpus size and exhaust service resources. This issue is fixed in version 2.0.0.

NVD references (4)

Sources

The tracker combines official, freely usable data sources – updated continuously:

This product uses data from the NVD API but is not endorsed or certified by the NVD. EPSS data © FIRST.org.

Affected? We can help.

We check whether your systems are affected, install updates and close the gap – continuously with the System Health Checker if you wish.