GHSA-8835-xmqj-354h: pgEdge Go LLM Library discloses provider API key fragments in error messages

Summary

The pgEdge Go LLM Library at versions 0.1.0 and 0.1.1 returns raw provider error text from its health handler, which can contain a partially masked fragment of the operator's LLM provider API key. GET /v1/health is unauthenticated unless the application sets an Authorize hook, so anyone who can reach the listening port can read the fragment.

Mitigation

Fixed in 0.2.0; upgrade to 0.2.0 or later, which redacts credentials from provider error messages. Rotate any LLM provider credential configured on an affected version where the proxy was reachable without an Authorize hook. Workarounds: set an Authorize hook on the proxy, or restrict network access to the listening port; neither removes the disclosure for a caller who reaches the endpoint.

Related items

• MediumOver 35,000 Grok-Built Apps Exposed Online, Some Leaking Confidential Data and PasswordsSimilar attack · The Hacker News
• MediumAI Coding Agents Leak Thousands of Secrets to GitHubSimilar attack · Dark Reading
• InfoPrivMark: Pixel-Space Steering against Privacy Leakage in Vision-Language ModelsSimilar attack · Arxiv (cs.RO + cs.CV security)
• MediumCVE-2026-121938: langchain4j-aideepin before 3.31.0 contains an insecure direct object reference vulnerability that allows authenticated…Similar attack · NVD/CVE Database
• InfoResidual Attribute Leakage Is an Interface: Auditing Privacy Evaluation After Text De-identificationSimilar attack · Arxiv (cs.CR + cs.CL + cs.LG)