CVE-2026-33075
FastGPT is an AI Agent building platform. In versions 4.14.8.3 and below, the fastgpt-preview-image.yml workflow is vulnerable to arbitrary code execution and secret exfiltration by any external contributor. It uses pull_request_target (which runs with access to repository secrets) but checks out code from the pull request author's fork, then builds and pushes Docker images using attacker-controlled Dockerfiles. This also enables a supply chain attack via the production container registry. A patch was not available at the time of publication.
Scoring
- Severity
- CRITICAL
- CVSS base score
- 9.4
- CVSS vector
- CVSS:4.0/AV:N/AC:L/AT:N/PR:L/UI:N/VC:H/VI:H/VA:H/SC:H/SI:H/SA:H
- EPSS probability
- 0.30%
- CWE
- CWE-494, CWE-829
- Published
- 2026-03-20
- Last modified
- 2026-03-20
Affected products
- labring FastGPT
Weakness type
Related vulnerabilities
- CVE-2026-79963 — Dell SCG 5.0 Appliance versions prior to 5.36.00.16 and Dell SCG 5.0 Application versions prior to...
- CVE-2026-62654 — A vulnerability has been identified in Reyrolle 7SR5 (All versions < V2.70). A special maintenance...
- CVE-2026-85427 — MOOS essential-moos through 10.0.1 pAntler Remote Code Execution via Unauthenticated MISSION_FILE
- CVE-2026-82021 — Hermes Agent 0.18.2 < 0.19.0 MCP Catalog Supply Chain RCE via Mutable Branch Reference
- CVE-2026-21810 — HCL BigFix Quantum Risk Analyzer is affected by a hardcoded external resource reference and downloading code without integrity checking
- CVE-2026-65081 — NVIDIA NemoClaw for Linux contains a vulnerability in its installation process, where an attacker...
- CVE-2026-65097 — NVIDIA NemoClaw for Linux contains a vulnerability in its installation scripts, where an attacker...
- CVE-2026-57910 — WatchGuard Agent improper authentication allows unauthenticated remote code execution