HighVulnerability
CVE-2026-54044: LibreChat credential exposure through agent action domain repointing
- Source
- NVD(opens in a new tab)
- Identifier
- CVE-2026-54044
- Published
- Record updated
- Known exploitation
- Not listed in the CISA Known Exploited Vulnerabilities catalog at the last check.
Summary
LibreChat versions before 0.8.7 let a collaborator with EDIT permission abuse POST /api/agents/actions/:agent_id in api/server/routes/agents/actions.js. The endpoint keeps the owner's stored metadata.api_key and OAuth client secrets while accepting editor-controlled metadata.domain and metadata.raw_spec, so the action can be repointed to an attacker-controlled host. A later tool invocation then decrypts and sends the retained owner credential to that destination.
Mitigation
Fixed in version 0.8.7.
Related items
- MediumCVE-2026-54026: LibreChat message injection via conversationId in request bodySimilar attack · NVD/CVE Database
- MediumCVE-2026-54046: LibreChat unauthenticated account enumeration via /api/user/verifySimilar attack · NVD/CVE Database
- MediumCVE-2026-54043: LibreChat is an enhanced ChatGPT clone that supports multiple AI providers. Prior to 0.8.7…Similar attack · NVD/CVE Database
- MediumCVE-2026-54042: LibreChat 2FA verification unlimited code guessing via temp token endpointSimilar attack · NVD/CVE Database
- HighCVE-2026-54039: LibreChat account takeover through registration and social login email fallbackSimilar attack · NVD/CVE Database