GHSA-8835-XMQJ-354H
Vulnerability from github – Published: 2026-11-12 19:32 – Updated: 2026-11-12 19:32Impact
An unauthenticated caller could read part of the operator's configured LLM provider API key.
The library's own health handler pings every configured provider using the operator's real credential and returns the resulting error text verbatim. Providers characteristically quote a partially masked form of the credential they rejected, so a failing ping published a fragment of the key.
The endpoint is unauthenticated by default. authorize() returns
true immediately when the consuming application sets no Authorize hook,
so GET /v1/health is reachable by anyone who can reach the listening
port unless the consumer opted in to authentication.
Who is affected: any application embedding pgedge-go-llm-lib at
0.1.0 or 0.1.1 that exposes the bundled HTTP proxy without setting an
Authorize hook. Both affected versions are released tags and resolve
through the public Go module proxy.
The disclosed material authenticates to a third-party provider account rather than to pgEdge software, which is why the severity carries Scope: Changed.
Evidence at v0.1.1: llm/proxy/handlers.go:296 (handleHealth
returning err.Error() in ProviderHealth.Error), llm/proxy/proxy.go:247
(authorize() defaulting to permit), and llm/proxy/proxy.go:350
(writeError writing the error with no redaction).
Patches
Fixed in 0.2.0. Upgrade to 0.2.0 or later, which redacts
credentials from provider error messages unconditionally. The fix adds
llm/internal/redact, with a follow-up anchoring word boundaries, and is
pinned by tests in that package.
Note the version. There is no 0.1.2 — it was never cut, and some
earlier material named it in error. 0.2.0 is the first fixed release.
Upgrading does not undo a disclosure that already happened. Rotate any
LLM provider credential configured on an affected version where the
proxy was reachable without an Authorize hook, particularly if a
provider was rejecting that key at the time.
Workarounds
Set an Authorize hook on the proxy, which closes the unauthenticated
route to the health endpoint. Restricting network access to the listening
port limits who can reach it. Neither removes the underlying disclosure
for a caller who does reach the endpoint.
{
"affected": [
{
"package": {
"ecosystem": "Go",
"name": "github.com/pgEdge/pgedge-go-llm-lib"
},
"ranges": [
{
"events": [
{
"introduced": "0"
},
{
"fixed": "0.1.2-0.20260729024727-7531a55d368b"
}
],
"type": "ECOSYSTEM"
}
]
}
],
"aliases": [
"CVE-2026-125328"
],
"database_specific": {
"cwe_ids": [
"CWE-209",
"CWE-522"
],
"github_reviewed": true,
"github_reviewed_at": "2026-11-12T19:32:36Z",
"nvd_published_at": "2026-11-10T19:21:10Z",
"severity": "MODERATE"
},
"details": "### Impact\n\nAn unauthenticated caller could read part of the operator\u0027s configured LLM provider API key.\n\nThe library\u0027s own health handler pings every configured provider using the operator\u0027s real credential and returns the resulting error text verbatim. Providers characteristically quote a partially masked form of the credential they rejected, so a failing ping published a fragment of\nthe key.\n\n**The endpoint is unauthenticated by default.** `authorize()` returns\ntrue immediately when the consuming application sets no `Authorize` hook,\nso `GET /v1/health` is reachable by anyone who can reach the listening\nport unless the consumer opted in to authentication.\n\n**Who is affected:** any application embedding pgedge-go-llm-lib at\n`0.1.0` or `0.1.1` that exposes the bundled HTTP proxy without setting an\n`Authorize` hook. Both affected versions are released tags and resolve\nthrough the public Go module proxy.\n\nThe disclosed material authenticates to a **third-party provider\naccount** rather than to pgEdge software, which is why the severity\ncarries Scope: Changed.\n\nEvidence at `v0.1.1`: `llm/proxy/handlers.go:296` (`handleHealth`\nreturning `err.Error()` in `ProviderHealth.Error`), `llm/proxy/proxy.go:247`\n(`authorize()` defaulting to permit), and `llm/proxy/proxy.go:350`\n(`writeError` writing the error with no redaction).\n\n### Patches\n\n**Fixed in `0.2.0`. Upgrade to `0.2.0` or later**, which redacts\ncredentials from provider error messages unconditionally. The fix adds\n`llm/internal/redact`, with a follow-up anchoring word boundaries, and is\npinned by tests in that package.\n\n**Note the version.** There is no `0.1.2` \u2014 it was never cut, and some\nearlier material named it in error. `0.2.0` is the first fixed release.\n\n**Upgrading does not undo a disclosure that already happened. Rotate any\nLLM provider credential configured on an affected version** where the\nproxy was reachable without an `Authorize` hook, particularly if a\nprovider was rejecting that key at the time.\n\n### Workarounds\n\nSet an `Authorize` hook on the proxy, which closes the unauthenticated\nroute to the health endpoint. Restricting network access to the listening\nport limits who can reach it. Neither removes the underlying disclosure\nfor a caller who does reach the endpoint.",
"id": "GHSA-8835-xmqj-354h",
"modified": "2026-11-12T19:32:36Z",
"published": "2026-11-12T19:32:36Z",
"references": [
{
"type": "WEB",
"url": "https://github.com/pgEdge/pgedge-go-llm-lib/security/advisories/GHSA-8835-xmqj-354h"
},
{
"type": "ADVISORY",
"url": "https://nvd.nist.gov/vuln/detail/CVE-2026-125328"
},
{
"type": "WEB",
"url": "https://github.com/pgEdge/pgedge-go-llm-lib/pull/26"
},
{
"type": "WEB",
"url": "https://github.com/pgEdge/pgedge-go-llm-lib/commit/7531a55d368b3a29cd173c792cc4609c4134b076"
},
{
"type": "WEB",
"url": "https://github.com/pgEdge/pgedge-go-llm-lib/commit/8bb9a6380d5f8db0124b51605759897034ad4da1"
},
{
"type": "PACKAGE",
"url": "https://github.com/pgEdge/pgedge-go-llm-lib"
},
{
"type": "WEB",
"url": "https://github.com/pgEdge/pgedge-go-llm-lib/releases/tag/v0.2.0"
}
],
"schema_version": "1.4.0",
"severity": [
{
"score": "CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:C/C:L/I:N/A:N",
"type": "CVSS_V3"
}
],
"summary": "pgEdge Go LLM Library discloses provider API key fragments in error messages"
}
Sightings
| Author | Source | Type | Date | Other |
|---|
Nomenclature
- Seen: The vulnerability was mentioned, discussed, or observed by the user.
- Confirmed: The vulnerability has been validated from an analyst's perspective.
- Published Proof of Concept: A public proof of concept is available for this vulnerability.
- Exploited: The vulnerability was observed as exploited by the user who reported the sighting.
- Patched: The vulnerability was observed as successfully patched by the user who reported the sighting.
- Not exploited: The vulnerability was not observed as exploited by the user who reported the sighting.
- Not confirmed: The user expressed doubt about the validity of the vulnerability.
- Not patched: The vulnerability was not observed as successfully patched by the user who reported the sighting.
The approach is described in our paper Mapping CVEs to MITRE ATT&CK Techniques: A Curated Gold-Set Classifier and the Limits of LLM-Assisted Label Expansion.
Browse all ATT&CK techniques and the vulnerabilities related to each.
Related by attack behaviour
Vulnerabilities whose description is nearest to this one in the vector space of the CIRCL/vulnerability-attack-technique-biencoder model. This is a similarity search over the bi-encoder space (plain cosine), not a classification, and it has no measured accuracy.