GHSA-rcjh-r59h-gq37 CVE-2026-48520 PYSEC-2026-244

Langflow: Unauthenticated Shareable Playground arbitrary local or S3 file read

Medium PyPI

Severity: 6.1 / 10; Medium · CVSS 3.1
Exploit chance (EPSS): 0.44%; Chance of exploitation in the next 30 days. Higher than 38.01% of scored vulnerabilities.
Known exploited: No; Not in CISA's catalog
Weakness: CWE-73 External Control of File Name or Path
Fix: 1.10.0; Upgrade langflow to this version or later

Description

Summary

The "Shareable Playground" (or "Public Flows" in code) contains a potential arbitrary file-read vulnerability, depending on the exact flow configuration used.

By making a flow public, public execution of the flow is allowed. The execution request can contain a list of files that gets read by Langflow and fed into the LLM. The files path can be any path supported by the storage - it can be either a local file or S3 path if supported by the local configuration

Details

Shareable Playground feature works by enabling the execution of workflows by unauthenticated users, by accessing a link. Specifically, it enables the route /api/v1/build_public_tmp to execute any public flow, given a public flow ID. This request contains a files field that can contain a list of files. The files get read in LCModelComponent._get_chat_result in a call to to_lc_message. A detailed stacktrace:

...
  File "/Users/ori/Work/research/langchain/langflow/src/backend/base/langflow/api/build.py", line 466, in build_vertices
    vertex_build_response: VertexBuildResponse = await _build_vertex(vertex_id, graph, event_manager)
  File "/Users/ori/Work/research/langchain/langflow/src/backend/base/langflow/api/build.py", line 324, in _build_vertex
    vertex_build_result = await graph.build_vertex(
  File "/Users/ori/Work/research/langchain/langflow/src/lfx/src/lfx/graph/graph/base.py", line 1563, in build_vertex
    await vertex.build(
  File "/Users/ori/Work/research/langchain/langflow/src/lfx/src/lfx/graph/vertex/base.py", line 770, in build
    await step(user_id=user_id, event_manager=event_manager, **kwargs)
  File "/Users/ori/Work/research/langchain/langflow/src/lfx/src/lfx/events/observability/lifecycle_events.py", line 95, in wrapper
    result = await observed_method(self, *args, **kwargs)
  File "/Users/ori/Work/research/langchain/langflow/src/lfx/src/lfx/graph/vertex/base.py", line 411, in _build
    await self._build_results(
  File "/Users/ori/Work/research/langchain/langflow/src/lfx/src/lfx/graph/vertex/base.py", line 640, in _build_results
    result = await initialize.loading.get_instance_results(
  File "/Users/ori/Work/research/langchain/langflow/src/lfx/src/lfx/interface/initialize/loading.py", line 76, in get_instance_results
    return await build_component(params=custom_params, custom_component=custom_component)
  File "/Users/ori/Work/research/langchain/langflow/src/lfx/src/lfx/interface/initialize/loading.py", line 299, in build_component
    build_results, artifacts = await custom_component.build_results()
  File "/Users/ori/Work/research/langchain/langflow/src/lfx/src/lfx/custom/custom_component/component.py", line 1136, in build_results
    return await self._build_with_tracing()
  File "/Users/ori/Work/research/langchain/langflow/src/lfx/src/lfx/custom/custom_component/component.py", line 1118, in _build_with_tracing
    results, artifacts = await self._build_results()
  File "/Users/ori/Work/research/langchain/langflow/src/lfx/src/lfx/custom/custom_component/component.py", line 1163, in _build_results
    result = await self._get_output_result(output)
  File "/Users/ori/Work/research/langchain/langflow/src/lfx/src/lfx/custom/custom_component/component.py", line 1238, in _get_output_result
    result = await method() if inspect.iscoroutinefunction(method) else await asyncio.to_thread(method)
  File "/Users/ori/Work/research/langchain/langflow/src/lfx/src/lfx/base/models/model.py", line 88, in text_response
    result = await self.get_chat_result(
  File "/Users/ori/Work/research/langchain/langflow/src/lfx/src/lfx/base/models/model.py", line 180, in get_chat_result
    return await self._get_chat_result(
  File "/Users/ori/Work/research/langchain/langflow/src/lfx/src/lfx/base/models/model.py", line 232, in _get_chat_result
    messages.append(input_value.to_lc_message(self.name))
  File "/Users/ori/Work/research/langchain/langflow/src/lfx/src/lfx/schema/message.py", line 184, in to_lc_message
    file_contents = self.get_file_content_dicts(model_name)
  File "/Users/ori/Work/research/langchain/langflow/src/lfx/src/lfx/schema/message.py", line 256, in get_file_content_dicts
    content_dicts.append(create_image_content_dict(file, None, model_name))
  File "/Users/ori/Work/research/langchain/langflow/src/lfx/src/lfx/utils/image.py", line 96, in create_image_content_dict
    ...

This triggers Langflow to feed the file into the LLM as an Image. Reading the files back depends on the specific LLM configuration.

PoC

Reproduction:

Create a new flow and add a Chat Input node to it
Share the flow ("Shareable Playground")
Access the public link with the browser developers tools open and execute the flow.
Find the /api/v1/build_public_tmp route and copy as cURL
Edit the files JSON field to point to any file.

Impact

Potential file read (local or S3) if shareable playground feature is used.

Ori Lahav Security Researcher @ Rubrik Inc.

Affected versions

PyPI

Package	Vulnerable	Fixed in
langflow	`< 1.10.0`	`1.10.0`

How to fix

Move to the fixed version or later with your package manager, update the lockfile, and run your tests. If the package only arrives through another dependency, upgrade that dependency first. If none of its releases pulls a fixed version yet, force the version with your package manager's override.

langflow PyPI → `1.10.0`

pip

pip install --upgrade "langflow>=1.10.0"

poetry

poetry add "langflow>=1.10.0"

uv

uv add "langflow>=1.10.0"

Severity breakdown

CVSS 3.1 6.1

CVSS:3.1/AV:N/AC:H/PR:N/UI:R/S:C/C:H/I:N/A:N

Attack vector: Network
Attack complexity: High
Privileges required: None
User interaction: Required
Scope: Changed
Confidentiality: High
Integrity: None
Availability: None

Weakness

CWE-73: External Control of File Name or Path

The product allows user input to control or influence paths or file names that are used in filesystem operations.

How to prevent it

Architecture and Design: When the set of filenames is limited or known, create a mapping from a set of fixed input values (such as numeric IDs) to the actual filenames, and reject all other inputs. For example, ID 1 could map to "inbox.txt" and ID 2 could map to "profile.txt". Features such as the ESAPI AccessReferenceMap provide this capability.
Architecture and Design, Operation: Run your code in a "jail" or similar sandbox environment that enforces strict boundaries between the process and the operating system. This may effectively restrict all access to files within a particular directory. Examples include the Unix chroot jail and AppArmor. In general, managed code may provide some protection. This may not be a feasible solution, and it only limits the impact to the operating system; the rest of your application may still be subject to compromise. Be careful to avoid CWE-243 and other weaknesses related to jails.
Architecture and Design: For any security checks that are performed on the client side, ensure that these checks are duplicated on the server side, in order to avoid CWE-602. Attackers can bypass the client-side checks by modifying values after the checks have been performed, or by changing the client to remove the client-side checks entirely. Then, these modified values would be submitted to the server.
Implementation: Assume all input is malicious. Use an "accept known good" input validation strategy, i.e., use a list of acceptable inputs that strictly conform to specifications. Reject any input that does not strictly conform to specifications, or transform it into something that does. When performing input validation, consider all potentially relevant properties, including length, type of input, the full range of acceptable values, missing or extra inputs, syntax, consistency across related fields, and conformance to business rules. As an example of business rule logic, "boat" may be syntactically valid because it only contains alphanumeric characters, but it is not valid if the input is only expected to contain colors such as "red" or "blue." Do not rely exclusively on looking for malicious or malformed inputs. This is likely to miss at least one undesirable input, especially if the code's environment changes. This can give attackers enough room to bypass the intended validation. However, denylists can be useful for detecting potential attacks or determining which inputs are so malformed that they should be rejected outright. When validating filenames, use stringent allowlists that limit the character set to be used. If feasible, only allow a single "." character in the filename to avoid weaknesses such as CWE-23, and exclude directory separators such as "/" to avoid CWE-36. Use a list of allowable file extensions, which will help to avoid CWE-434. Do not rely exclusively on a filtering mechanism that removes potentially dangerous characters. This is equivalent to a denylist, which may be incomplete (CWE-184). For example, filtering "/" is insufficient protection if the filesystem also supports the use of "\" as a directory separator. Another possible error could occur when the filtering is applied in a way that still produces dangerous data (CWE-182). For example, if "../" sequences are removed from the ".../...//" string in a sequential fashion, two instances of "../" would b
Implementation: Use a built-in path canonicalization function (such as realpath() in C) that produces the canonical version of the pathname, which effectively removes ".." sequences and symbolic links (CWE-23, CWE-59).
Installation, Operation: Use OS-level permissions and run as a low-privileged user to limit the scope of any successful attack.
Operation, Implementation: If you are using PHP, configure your application so that it does not use register_globals. During implementation, develop your application so that it does not rely on this feature, but be wary of implementing a register_globals emulation that is subject to weaknesses such as CWE-95, CWE-621, and similar issues.
Testing: Use tools and techniques that require manual (human) analysis, such as penetration testing, threat modeling, and interactive tools that allow the tester to record and modify an active session. These may be more effective than strictly automated techniques. This is especially the case with weaknesses that are related to design and business rules.

References

Advisories

nvd.nist.gov/vuln/detail/CVE-2026-48520

Packages

github.com/langflow-ai/langflow

Timeline

Jun 16, 2026 Published
Oct 26, 2026 Last updated by the source
Oct 31, 2026 Latest EPSS score

Advisory	Severity	Why it's related
GHSA-qrpv-q767-xqq2 Langflow: IDOR Vulnerability in `/api/v1/responses` Endpoint Allows Authenticated Attackers to Access Another User's Flow	High Known exploited	Also affects langflow
GHSA-vwmf-pq79-vjvx Unauthenticated Remote Code Execution in Langflow via Public Flow Build Endpoint	Critical Known exploited	Also affects langflow
GHSA-g22f-v6f7-2hrh Langflow affected by Remote Code Execution via validate_code() exec()	High Known exploited	Also affects langflow
GHSA-577h-p2hh-v4mv Langflow CORS misconfiguration enables Account Takeover and RCE	Critical Known exploited	Also affects langflow
GHSA-rvqx-wpfh-mfx7 Langflow Unauth RCE	Critical Known exploited	Also affects langflow
GHSA-j8f7-x8jm-wmm4 Langflow: SSRF in URL-taking components (protection disabled by default / warn-only, not applied to RSS, SearXNG, Web Search, Home Assistant, Glean, Docling)	Medium	Also affects langflow
GHSA-4f6c-2vvp-gw82 Langflow: IP Spoofing Bypass via `X-Forwarded-For` Allowing Remote Configuration Write	High	Also affects langflow
GHSA-4hmc-cfm3-w43c Langflow has Authenticated Cross-Project File Disclosure via Unscoped MCP Resource Handlers	High	Also affects langflow

Is this in your code?

MergeHarbor checks every repository's dependencies against this database, ranks what matters by exploitability and reachability, routes it to the owning team and opens the fix pull request.

Check your repos with MergeHarbor