/

CVE-2026-79986

Severity:

HIGH RISK

(7.4)

Path Traversal (CWE-22) in Apache Camel Azure Files

CVE-2026-79986

Modified

Vulnerability type

Category:

Path Traversal

CWE:

CWE-22

Impact:

Unauthorized remote file access and manipulation

Severity:

HIGH RISK

(7.4)

CVSS version:

3.1

Catch this CVE while code is being written

Detect vulnerable code directly in pull requests, so CVEs are fixed before merge.

Summary

Apache Camel's Azure Files component may construct remote file paths from directory-listing names without normalizing them or verifying that they remain inside the configured polling directory. If a remote service or an on-path attacker supplies a listing entry containing path components, Camel may read, delete, or rename files outside the intended directory using the endpoint's credentials.

Apache Camel's Azure Files component may construct remote file paths from directory-listing names without normalizing them or verifying that they remain inside the configured polling directory. If a remote service or an on-path attacker supplies a listing entry containing path components, Camel may read, delete, or rename files outside the intended directory using the endpoint's credentials.

Affected context

The issue occurs in Apache Camel remote-file consumers using the Azure Files component when a listing entry is treated as more than a single path segment. The consumer concatenates the configured directory with the reported name, while local work-directory protections and the jailStartingDirectory setting do not constrain this remote operand. Azure Files services may not normally return such names, but exposure can arise with malicious, compromised, multi-tenant, virtualized, or tampered remote services.

The issue occurs in Apache Camel remote-file consumers using the Azure Files component when a listing entry is treated as more than a single path segment. The consumer concatenates the configured directory with the reported name, while local work-directory protections and the jailStartingDirectory setting do not constrain this remote operand. Azure Files services may not normally return such names, but exposure can arise with malicious, compromised, multi-tenant, virtualized, or tampered remote services.

Exploit preconditions

  • Attack surface: Apache Camel Azure Files remote-file consumer
  • Required access: Ability to influence the remote directory listing, including through a malicious or compromised service or an on-path attacker able to alter an unauthenticated FTP-like listing channel where applicable
  • Trigger point: Processing a listed file during polling
  • Additional conditions: Retrieval is available in normal consumer operation; deletion or renaming also requires options such as delete, move, preMove, moveFailed, or rename-based read locking
  • Why exploitation is possible: The reported name is concatenated into the remote path without lexical normalization or containment validation
  • Attack surface: Apache Camel Azure Files remote-file consumer
  • Required access: Ability to influence the remote directory listing, including through a malicious or compromised service or an on-path attacker able to alter an unauthenticated FTP-like listing channel where applicable
  • Trigger point: Processing a listed file during polling
  • Additional conditions: Retrieval is available in normal consumer operation; deletion or renaming also requires options such as delete, move, preMove, moveFailed, or rename-based read locking
  • Why exploitation is possible: The reported name is concatenated into the remote path without lexical normalization or containment validation

Why this happens

The underlying bug is improper limitation of a pathname to a restricted directory, corresponding to CWE-22: Path Traversal. The consumer trusted a server-provided listing name, combined it with the configured directory, and used the resulting path for later remote operations without verifying that the resolved path remained beneath the intended directory. This insecure pattern allows path components in metadata to escape the configured logical boundary.

The underlying bug is improper limitation of a pathname to a restricted directory, corresponding to CWE-22: Path Traversal. The consumer trusted a server-provided listing name, combined it with the configured directory, and used the resulting path for later remote operations without verifying that the resolved path remained beneath the intended directory. This insecure pattern allows path components in metadata to escape the configured logical boundary.

Potential impact

  • Camel may read files outside the configured polling directory.
  • Camel may delete files outside the intended directory when deletion is enabled.
  • Camel may rename or move files outside the intended directory when disposition or read-lock options permit it.
  • Sensitive remote data may be disclosed through route processing or downstream storage.
  • Files accessible to the endpoint account could be altered or removed.
  • Camel may read files outside the configured polling directory.
  • Camel may delete files outside the intended directory when deletion is enabled.
  • Camel may rename or move files outside the intended directory when disposition or read-lock options permit it.
  • Sensitive remote data may be disclosed through route processing or downstream storage.
  • Files accessible to the endpoint account could be altered or removed.

General mitigation guidance

Mitigation focuses on enforcing the directory boundary at the point where remote paths are constructed and before every remote file operation.

Common defensive patterns include:

  • Normalize remote path components before joining them with the configured directory.
  • Reject names containing parent-directory components, absolute-path indicators, separators, or other unexpected structure when only a single filename is valid.
  • Canonicalize or resolve the candidate path and require verified containment beneath the configured directory.
  • Apply the same containment policy consistently to retrieval, deletion, renaming, failed-move handling, and read-lock operations.
  • Treat remote listings as untrusted input rather than route-controlled data.
  • Use endpoint credentials with only the minimum remote permissions required, preferably separating read-only consumers from workflows that delete or rename files.

Mitigation focuses on enforcing the directory boundary at the point where remote paths are constructed and before every remote file operation.

Common defensive patterns include:

  • Normalize remote path components before joining them with the configured directory.
  • Reject names containing parent-directory components, absolute-path indicators, separators, or other unexpected structure when only a single filename is valid.
  • Canonicalize or resolve the candidate path and require verified containment beneath the configured directory.
  • Apply the same containment policy consistently to retrieval, deletion, renaming, failed-move handling, and read-lock operations.
  • Treat remote listings as untrusted input rather than route-controlled data.
  • Use endpoint credentials with only the minimum remote permissions required, preferably separating read-only consumers from workflows that delete or rename files.

How teams detect issues like this

Detection can combine source review, automated analysis, and runtime policy controls:

  • Search for remote path construction based on directory-listing names without normalization or containment checks.
  • Review calls to path-joining, concatenation, retrieval, deletion, and rename APIs for untrusted filename input.
  • Flag assumptions that a remote listing always contains a single path segment.
  • Add tests covering parent-directory components, absolute paths, separators, encoded traversal forms, and unexpected virtual-filesystem syntax.
  • Monitor consumer activity for access, deletion, or rename operations outside approved remote directory prefixes.
  • Use static analysis or automated code-review rules that identify filesystem or remote-file operations lacking boundary validation.

Detection can combine source review, automated analysis, and runtime policy controls:

  • Search for remote path construction based on directory-listing names without normalization or containment checks.
  • Review calls to path-joining, concatenation, retrieval, deletion, and rename APIs for untrusted filename input.
  • Flag assumptions that a remote listing always contains a single path segment.
  • Add tests covering parent-directory components, absolute paths, separators, encoded traversal forms, and unexpected virtual-filesystem syntax.
  • Monitor consumer activity for access, deletion, or rename operations outside approved remote directory prefixes.
  • Use static analysis or automated code-review rules that identify filesystem or remote-file operations lacking boundary validation.

Sources

SourceLink
camel.apache.orgcamel.apache.org/security/CVE-2026-79986.html
github.comgithub.com/apache/camel/pull/25730
github.comgithub.com/apache/camel/pull/25745
github.comgithub.com/apache/camel/pull/25746
github.comgithub.com/apache/camel/pull/25747
issues.apache.orgissues.apache.org/jira/browse/CAMEL-24487
github.comgithub.com/oscerd/CVE-2026-79986

Frequently asked

questions

It may be, when an attacker can influence the remote directory listing or alter the listing data in transit. A malicious, compromised, or unusual remote service may also provide a crafted listing entry.

Not necessarily. The attacker needs influence over the listing source or its transport, while Camel supplies the endpoint credentials for the resulting remote operation.

Treat listing names as untrusted, normalize them, and enforce that every resolved remote path remains within the configured directory before retrieval, deletion, or renaming.

Use strict single-segment filename validation where appropriate, apply least-privilege endpoint accounts, test traversal variants, and enforce automated review policies for remote path construction.

It may be, when an attacker can influence the remote directory listing or alter the listing data in transit. A malicious, compromised, or unusual remote service may also provide a crafted listing entry.

Not necessarily. The attacker needs influence over the listing source or its transport, while Camel supplies the endpoint credentials for the resulting remote operation.

Treat listing names as untrusted, normalize them, and enforce that every resolved remote path remains within the configured directory before retrieval, deletion, or renaming.

Use strict single-segment filename validation where appropriate, apply least-privilege endpoint accounts, test traversal variants, and enforce automated review policies for remote path construction.

Find out what's already
exploitable in your codebase.