Executive brief
Flowise, an open-source tool for building LLM applications, contains a security flaw where it uses easily guessable default passwords (secrets) for its login tokens. If an administrator does not manually change these settings during setup, an attacker can use these known defaults to create fake login credentials. This allows an attacker to impersonate any user, including administrators, potentially gaining full control over the application and its data.
Technical details
Flowise (npm package 'flowise') contains a vulnerability where JWT secrets, audiences, and issuers fall back to weak, hardcoded strings if environment variables (JWT_AUTH_TOKEN_SECRET, JWT_REFRESH_TOKEN_SECRET, etc.) are not defined. Specifically, in 'packages/server/src/enterprise/middleware/passport/index.ts', the application defaults to values like 'auth_token' and 'refresh_token'. Because these defaults are public knowledge, an attacker can forge valid JSON Web Tokens (JWTs) to impersonate any user account, including administrators. Additionally, a design flaw exists where the refresh token secret falls back to the auth token secret if the former is unset, further reducing the cryptographic entropy of the system. This issue is resolved in version 3.1.0.
Affected products
- FlowiseAI flowise <= 3.0.13
Timeline
- 2026-04-15: disclosed
- 2026-04-16: advisory: GitHub reviewed advisory published
- 2026-04-16: patched: Fixed in version 3.1.0
References
- https://api.github.com/users/kolega-ai-dev
- https://github.com/kolega-ai-dev
- https://api.github.com/users/kolega-ai-dev/gists%7B/gist_id%7D
- https://api.github.com/users/kolega-ai-dev/repos
- https://avatars.githubusercontent.com/u/250560052?v=4
- https://api.github.com/users/kolega-ai-dev/events%7B/privacy%7D