CVE-2026-70477 FlowiseAI CVE debrief

Who should care

Defenders responsible for Flowise deployments should assess exposure and prioritize remediation. This vulnerability allows for arbitrary code execution in the context of the service account, which can lead to a compromise of the service and potentially the underlying system.

Why it matters

CVE-2026-70477 is a critical vulnerability in Flowise that allows for arbitrary code execution. Defenders should prioritize remediation to prevent potential compromise of the service and underlying system.

• Arbitrary code execution in the context of the service account.
• Potential compromise of the Flowise service and underlying system.
• Bypass of the blocklist validator.
• Execution of malicious Python scripts in an unsandboxed environment.

Technical summary

A prompt injection vulnerability exists in Flowise, a drag & drop user interface to build a customized large language model flow. Prior to version 3.1.3, an attacker can inject a malicious Python script that bypasses the blocklist validator and executes in an unsandboxed Pyodide environment. The vulnerability is due to the use of untrusted data to construct an LLM prompt in the run method of the CSV_Agents class.

Defensive priority

High

Recommended defensive actions

• Review and apply the vendor's patch to upgrade to Flowise version 3.1.3 or later.
• Restrict access to the Flowise service to trusted users and networks.
• Monitor the Flowise service for suspicious activity and implement logging and alerting.
• Verify the integrity of the Flowise environment and configurations.
• Conduct a thorough review of the Flowise service to identify potential exposure.
• Implement compensating controls for exposed systems while remediation is scheduled and verified.
• Track exceptions, retest remediated assets, and close the item only after evidence is documented.

Evidence notes

The CVE record and NVD entry provide details on the vulnerability. The vendor, FlowiseAI, has released a fix in version 3.1.3. However, the specifics of the vulnerability, such as the exact impact and affected configurations, require verification from official sources.

Sources and references

Methodology and review provenance

AI-assisted synthesis based on stored public vulnerability evidence. System validation, approval state, and publication status do not by themselves establish human review of this revision. PatchSiren helps prioritize defensive review and does not prove exposure or remediation on any system.