GHSA-8C4J-F57C-35CF

Vulnerability from github – Published: 2026-03-27 19:36 – Updated: 2026-03-27 21:52
VLAI
Summary
Langflow: Authenticated Users Can Read, Modify, and Delete Any Flow via Missing Ownership Check
Details

Vulnerability

IDOR in GET/PATCH/DELETE /api/v1/flow/{flow_id}

The _read_flow helper in src/backend/base/langflow/api/v1/flows.py branched on the AUTO_LOGIN setting to decide whether to filter by user_id. When AUTO_LOGIN was False (i.e., authentication was enabled), neither branch enforced an ownership check — the query returned any flow matching the given UUID regardless of who owned it.

This exposed any authenticated user to:

  • Read any other user's flow, including embedded plaintext API keys
  • Modify the logic of another user's AI agents
  • Delete flows belonging to other users

The vulnerability was introduced by the conditional logic that was meant to accommodate public/example flows (those with user_id = NULL) under auto-login mode, but inadvertently left the authenticated path without an ownership filter.

Fix (PR #8956)

The fix removes the AUTO_LOGIN conditional entirely and unconditionally scopes the query to the requesting user:

-    auth_settings = settings_service.auth_settings
-    stmt = select(Flow).where(Flow.id == flow_id)
-    if auth_settings.AUTO_LOGIN:
-        stmt = stmt.where(
-            (Flow.user_id == user_id) | (Flow.user_id == None)  # noqa: E711
-        )
+    stmt = select(Flow).where(Flow.id == flow_id).where(Flow.user_id == user_id)

All three operations — read, update, and delete — route through _read_flow, so the single change covers the full attack surface. A cross-user isolation test (test_read_flows_user_isolation) was added to prevent regression.

Acknowledgements

Langflow thanks the security researcher who responsibly disclosed this vulnerability:

Severity
8.7 (High)
CVSS:4.0/AV:N/AC:L/AT:N/PR:L/UI:N/VC:H/VI:H/VA:H/SC:N/SI:N/SA:N
Show details on source website
To clipboard 

{
  "affected": [
    {
      "database_specific": {
        "last_known_affected_version_range": "\u003c= 1.5.0"
      },
      "package": {
        "ecosystem": "PyPI",
        "name": "langflow"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "0"
            },
            {
              "fixed": "1.5.1"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    },
    {
      "database_specific": {
        "last_known_affected_version_range": "\u003c= 0.5.0"
      },
      "package": {
        "ecosystem": "PyPI",
        "name": "langflow-base"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "0"
            },
            {
              "fixed": "0.5.1"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    }
  ],
  "aliases": [
    "CVE-2026-34046"
  ],
  "database_specific": {
    "cwe_ids": [
      "CWE-639",
      "CWE-862"
    ],
    "github_reviewed": true,
    "github_reviewed_at": "2026-03-27T19:36:23Z",
    "nvd_published_at": "2026-03-27T21:17:27Z",
    "severity": "HIGH"
  },
  "details": "## Vulnerability\n\n### IDOR in `GET/PATCH/DELETE /api/v1/flow/{flow_id}`\n\nThe `_read_flow` helper in `src/backend/base/langflow/api/v1/flows.py` branched on the `AUTO_LOGIN` setting to decide whether to filter by `user_id`. When `AUTO_LOGIN` was `False` (i.e., authentication was enabled), neither branch enforced an ownership check \u2014 the query returned any flow matching the given UUID regardless of who owned it.\n\nThis exposed any authenticated user to:\n\n- **Read** any other user\u0027s flow, including embedded plaintext API keys\n- **Modify** the logic of another user\u0027s AI agents\n- **Delete** flows belonging to other users\n\nThe vulnerability was introduced by the conditional logic that was meant to accommodate public/example flows (those with `user_id = NULL`) under auto-login mode, but inadvertently left the authenticated path without an ownership filter.\n\n---\n\n## Fix (PR #8956)\n\nThe fix removes the `AUTO_LOGIN` conditional entirely and unconditionally scopes the query to the requesting user:\n\n```diff\n-    auth_settings = settings_service.auth_settings\n-    stmt = select(Flow).where(Flow.id == flow_id)\n-    if auth_settings.AUTO_LOGIN:\n-        stmt = stmt.where(\n-            (Flow.user_id == user_id) | (Flow.user_id == None)  # noqa: E711\n-        )\n+    stmt = select(Flow).where(Flow.id == flow_id).where(Flow.user_id == user_id)\n```\n\nAll three operations \u2014 read, update, and delete \u2014 route through `_read_flow`, so the single change covers the full attack surface. A cross-user isolation test (`test_read_flows_user_isolation`) was added to prevent regression.\n\n---\n\n## Acknowledgements\n\nLangflow thanks the security researcher who responsibly disclosed this vulnerability:\n\n- **[@chximn-dt](https://github.com/chximn-dt)**",
  "id": "GHSA-8c4j-f57c-35cf",
  "modified": "2026-03-27T21:52:39Z",
  "published": "2026-03-27T19:36:23Z",
  "references": [
    {
      "type": "WEB",
      "url": "https://github.com/langflow-ai/langflow/security/advisories/GHSA-8c4j-f57c-35cf"
    },
    {
      "type": "ADVISORY",
      "url": "https://nvd.nist.gov/vuln/detail/CVE-2026-34046"
    },
    {
      "type": "WEB",
      "url": "https://github.com/langflow-ai/langflow/pull/8956"
    },
    {
      "type": "PACKAGE",
      "url": "https://github.com/langflow-ai/langflow"
    }
  ],
  "schema_version": "1.4.0",
  "severity": [
    {
      "score": "CVSS:4.0/AV:N/AC:L/AT:N/PR:L/UI:N/VC:H/VI:H/VA:H/SC:N/SI:N/SA:N",
      "type": "CVSS_V4"
    }
  ],
  "summary": "Langflow: Authenticated Users Can Read, Modify, and Delete Any Flow via Missing Ownership Check"
}



Log in or create an account to share your comment.




Tags
Taxonomy of the tags.





Forecast uses a logistic model when the trend is rising, or an exponential decay model when the trend is falling. Fitted via linearized least squares.

Sightings

Author Source Type Date Other

Nomenclature

  • Seen: The vulnerability was mentioned, discussed, or observed by the user.
  • Confirmed: The vulnerability has been validated from an analyst's perspective.
  • Published Proof of Concept: A public proof of concept is available for this vulnerability.
  • Exploited: The vulnerability was observed as exploited by the user who reported the sighting.
  • Patched: The vulnerability was observed as successfully patched by the user who reported the sighting.
  • Not exploited: The vulnerability was not observed as exploited by the user who reported the sighting.
  • Not confirmed: The user expressed doubt about the validity of the vulnerability.
  • Not patched: The vulnerability was not observed as successfully patched by the user who reported the sighting.



Related by attack behaviour

Vulnerabilities whose description is nearest to this one in the vector space of the CIRCL/vulnerability-attack-technique-biencoder model. This is a similarity search over the bi-encoder space (plain cosine), not a classification, and it has no measured accuracy.