Authenticated command execution via evaluated sandbox output
Published Dec 31, 2025 · Updated Dec 31, 2025
Code injection in RAGFlow before 0.23.0 allows remote authenticated users to execute system commands through the Canvas CodeExec component. agent/tools/code_exec.py passes sandbox-controlled stdout to Python eval(), so a printed expression runs in the server host process instead of remaining sandboxed. A normal login account with access to create and run a Canvas CodeExec node is sufficient, and successful exploitation grants host-process command execution.
Summary
What happened
Code injection in RAGFlow before 0.23.0 allows remote authenticated users to execute system commands through the Canvas CodeExec component. agent/tools/code_exec.py passes sandbox-controlled stdout to Python eval(), so a printed expression runs in the server host process instead of remaining sandboxed. A normal login account with access to create and run a Canvas CodeExec node is sufficient, and successful exploitation grants host-process command execution.
The record
- CVE
- CVE-2025-68700
- Published
- Dec 31, 2025
- Updated
- Dec 31, 2025
- Vendor
- InfiniFlow
- Product
- RAGFlow
- Classifications
- CWE-78, T1059.006
- Attack vector
- network
- Privileges
- authenticated
Timeline
How it unfolded
- Dec 31, 2025CVE publishedPublication date reported by the CVE source.
- Dec 31, 2025Record updatedLatest update available in the CVE record.
Exploitability
Present is not the same as exploitable
Compare your product and version with the public record. A matching version still requires validation against your environment.
Is a vulnerable build present?
Compare these published version ranges with your installed build and any vendor patches.
- Affected versionversion=< 0.23.0
What conditions does exploitation require?
What is affected?
Published CVSS scores
CVSS describes severity. EPSS estimates exploitation probability.
Attacks
What attackers are doing with it
Daily unique IPs observed by Shadowserver honeypots for known exploited vulnerabilities (KEVs). Missing observations do not establish an absence of attacks.
Weakness, pattern, technique
Public exploit references
- Canvas CodeExec host-command proof of conceptfunctional · demonstrated
Labels summarize the accepted research assessment. They do not indicate a test against your environment.
Technologies
Your stack
See the directory against your own environment.
Your stack
Check the software in your environment
Book a demo to see how Hinoki identifies affected software and validates exploitability in your environment.
Book a demo