Description

Flowise before 3.1.3 contains an incomplete credential redaction vulnerability in the GET /api/v1/credentials/:id endpoint that returns decrypted secrets in plaintext. Authenticated users with credentials:view permission can retrieve sensitive data including database connection URLs with embedded passwords, cloud service account JSON with private keys, and API keys by calling this endpoint.

Executive Summary

Technical Details

Mitigations

References

Additional Metadata

{
  "score": "6.5",
  "pubDate": "2026-08-13T12:17:24.750Z",
  "pubdate": "2026-08-13T12:17:24.750Z",
  "executiveSummary": "An incomplete credential redaction vulnerability exists in Flowise before version 3.1.3 within the GET /api/v1/credentials/:id endpoint.\nThe vulnerability allows authenticated users possessing the credentials:view permission to retrieve sensitive decrypted secrets in plaintext.\nAffected systems expose high-value assets including database connection URLs with embedded passwords, cloud service account JSON objects containing private cryptographic keys, and third-party API keys.\nThe risk implications are severe, as unauthorized exposure of administrative credentials and private keys can lead to lateral movement, unauthorized data access, and complete compromise of integrated external cloud and database services.\nAttacker capabilities are constrained to authenticated users with specific viewing privileges, requiring active session access or compromised credentials of a user with the credentials:view role.\nExploitation requires interacting directly with the vulnerable API endpoint to extract sensitive material meant to be redacted or masked during transmission.",
  "technicalDetails": "The vulnerability resides in the credential management subsystem of Flowise, specifically affecting the GET /api/v1/credentials/:id API route.\nThe root cause stems from an incomplete credential redaction implementation where backend logic fails to properly sanitize or mask sensitive fields prior to serializing the HTTP response object for retrieval requests.\nAffected versions include all Flowise deployments prior to version 3.1.3.\nThe vulnerability requires network exposure of the Flowise application interface, authentication credentials for an account, and the presence of the specific credentials:view permission assigned to the attacking principal.\nThe attack flow proceeds as follows: 1) The authenticated user crafts an HTTP GET request targeting the vulnerable endpoint GET /api/v1/credentials/:id, substituting :id with a valid target credential identifier. 2) The request is processed by the Flowise backend service, which queries the internal database or credential store. 3) Instead of applying robust redaction filters to sensitive properties, the application retrieves the plaintext secrets or successfully decrypts stored payloads (such as database connection strings containing plaintext passwords, cloud service account JSON structures housing private keys, and API keys) and populates the response body. 4) The server responds with an HTTP payload containing the fully decrypted, unmasked sensitive secrets in plaintext format.\nThe post-exploitation impact includes the aggregation of high-privilege secrets that can be leveraged to pivot into connected relational databases, execute remote administrative actions on cloud infrastructure utilizing exposed service account private keys, and abuse external API integrations associated with the victimized Flowise instance."
}