File write and command execution via LLM tools
Published Jul 11, 2026 · Updated Jul 11, 2026
Path traversal and command injection in Mervin Praison PraisonAI before 4.6.78 allow remote authenticated users to write files and execute commands. AICoder joins an LLM-supplied absolute path to its working directory, opens it for writing, and passes LLM-generated command text to a subprocess without validation, an allowlist, or a sandbox. Access to the chat interface is required; in root-running Docker deployments, injected prompts can overwrite privileged files and execute commands as root.
Summary
What happened
Path traversal and command injection in Mervin Praison PraisonAI before 4.6.78 allow remote authenticated users to write files and execute commands. AICoder joins an LLM-supplied absolute path to its working directory, opens it for writing, and passes LLM-generated command text to a subprocess without validation, an allowlist, or a sandbox. Access to the chat interface is required; in root-running Docker deployments, injected prompts can overwrite privileged files and execute commands as root.
The record
- CVE
- CVE-2026-61445
- Published
- Jul 11, 2026
- Updated
- Jul 11, 2026
- Vendor
- Mervin Praison
- Product
- PraisonAI
- Classifications
- CWE-22, CWE-78, T1059
- Attack vector
- network
- Privileges
- authenticated
Timeline
How it unfolded
- Jul 11, 2026CVE publishedPublication date reported by the CVE source.
- Jul 11, 2026Record 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 <4.6.78
What conditions does exploitation require?
What is affected?
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
- AICoder prompt-injection proof of conceptproof of concept · 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