Summary
ArcadeDB does not sufficiently sanitize database names submitted through the server database-management API. An authenticated root user may abuse directory traversal sequences to create databases or recursively delete files and directories outside the configured database directory.
Affected context
The vulnerability occurs in ArcadeDB deployments that expose the POST /api/v1/server endpoint for database creation and deletion and permit authenticated root users to invoke those commands. It becomes risky when database names are incorporated into filesystem paths without canonicalization and boundary checks, allowing the resulting path to escape the intended database directory.
Exploit preconditions
- Attack surface: The
POST /api/v1/serverdatabase-management endpoint - Required access: Authentication as a user with root-level permissions
- Trigger point: Database creation or deletion commands that process a
caller-controlled database name
- Why exploitation is possible: Traversal sequences in the database name
can influence filesystem path resolution, and the server does not reliably
constrain the resolved path to the configured database directory
Why this happens
The underlying bug is improper limitation of a user-controlled pathname, corresponding to CWE-22, Path Traversal. Database names are treated as safe path components without consistently rejecting traversal elements, canonicalizing the resulting path, and verifying that it remains beneath the permitted database root. As a result, filesystem operations intended for database storage may be redirected to attacker-selected locations accessible to the server process.
Potential impact
- An attacker may create database-related files in unintended filesystem locations.
- An attacker may delete files or recursively remove directories that the
server process can access.
- The vulnerability could cause data loss or service interruption.
- It may enable tampering with application configuration or runtime data.
- It could increase the impact of a separate compromise involving the database
server or its surrounding environment.
General mitigation guidance
Mitigation generally focuses on treating database names as restricted
identifiers rather than filesystem paths.
Common defensive patterns include:
- Accepting only a narrow, documented character set for database names.
- Rejecting traversal markers, separators, absolute paths, and encoded
equivalents before path construction.
- Canonicalizing candidate paths and enforcing that they remain beneath the
configured database directory.
- Using filesystem APIs that safely join and constrain paths where available.
- Separating database-management permissions from broader administrative
privileges and applying least privilege to the server process.
- Applying additional confirmation and policy controls to recursive deletion
operations.
How teams detect issues like this
Detection approaches may include:
- Static analysis rules for concatenating request parameters with filesystem
paths without canonicalization or containment checks.
- Code review checks for database or resource names used directly in create,
delete, move, or recursive-removal operations.
- API gateway or application policies that reject path separators, traversal
patterns, and encoded variants in database-name parameters.
- Audit monitoring for unusual database names, operations targeting unexpected
locations, and recursive deletion requests.
- Automated security testing that verifies resolved paths cannot escape the
configured storage directory.

