Critical severity · published Aug 13, 2026
CVE-2026-73487
Flowise before 3.1.3 Prompt Injection RCE via CSV Agent
9.8
CVSS
1%
EPSS
No
On KEV
Description
Flowise before 3.1.3 contains a regex-based Python code validator bypass in CSV and Airtable Agent nodes that allows unauthenticated attackers to inject malicious code via prompt injection. Attackers can exploit unblocked pandas functions like pd.read_json() to exfiltrate datasets, perform SSRF against internal services, or achieve code execution through the unauthenticated prediction API.
CVSS vector
CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H
Attack vector
Network
Complexity
Low
Privileges
None
User interaction
None
Confidentiality
High
Integrity
High
Availability
High
Scope
Unchanged
Exploitation
Is it being used?
Exploit prediction (EPSS)1.0% · higher than 61% of CVEs
Exploited in the wildNo evidence yet
CISA KEVNot listed
Ransomware campaign useNot known
Affected products
Which versions
FL
Flowise · FlowiseAI
before 3.1.3; before 3.1.3 · source: vendor, NVD
Affected product
FL
Flowise
FlowiseAI
More from Flowise
10.0
CVE-2026-70478
Flowise: Unauthenticated OAuth2 token refresh endpoint returns access tokens — enables token theft for any connected service
10.0
CVE-2025-59528
Flowise has Remote Code Execution vulnerability
9.9
CVE-2026-73602
Flowise before 3.1.3 Sandbox Escape to RCE
9.9
CVE-2026-67622
Flowise 3.1.4 IDOR in OpenAI Assistants Integration
9.9
CVE-2026-56274
Flowise - Remote Code Execution via MCP Security Bypass in validateCommandFlags and validateArgsForLocalFileAccess