Vanna prompt injection code execution
CVE-2024-5565 · GHSA-7735-w2jp-gvg6 · published 2024-05-31 · updated 2024-11-25
- Severity
- Critical
- CVSS score
- 8.1
- Chance of exploitation (EPSS)
- 15.2%
- Fixed in
- No fix yet
- Exploitation Public proof of concept
- Automatable No
- Technical impact Total
- CISA SSVC
The Vanna library uses a prompt function to present the user with visualized results, it is possible to alter the prompt using prompt injection and run arbitrary Python code instead of the intended visualization code. Specifically - allowing external input to the library’s “ask” method with "visualize" set to True (default behavior) leads to remote code execution.
No fixed release is listed yet for vanna. Follow the source advisory for updates, and until one ships:
General guidance from ShieldPi, not a vendor fix
- Command or code injection. Remove or disable the affected server or tool, or run it in a sandbox with no secrets in its environment.
- Prompt injection or tool poisoning. Treat what its tools return as untrusted input, and require a person's approval before it sends data out or takes any other outbound action.Watchtower Guard can check tool results against its prompt-injection patterns when you send them as user_input; reworded, translated or encoded injections can pass them. Docs →
| Package | Registry | Affected | Fixed in |
|---|---|---|---|
| vanna | PyPI | <= 0.5.5 | No fix yet |
- CriticalUnauthenticated Remote Code Execution in Langflow via Public Flow Build EndpointlangflowExploited
- CriticalRay is vulnerable to Critical RCE via Safari & Firefox Browsers through DNS Rebinding AttackrayExploited
- CriticalLangflow Unauth RCElangflow, langflow-baseExploited
- HighLiteLLM: Authenticated command execution via MCP stdio test endpointslitellmExploited
- CriticalMLflow vulnerable to remote code execution on unauthenticated servers via third-party scorer deserialization importing caller-placed artifacmlflow
- CriticalAgno: Arbitrary code execution via `eval()` on `field_type` in `FunctionCall`agno
The impact, patch and workaround text comes from the advisory, shortened (GitHub Advisory Database, CC BY 4.0). Proof-of-concept and reproduction sections and code are left out; the original advisory has them.