An MLflow tracking API used an untrusted model identifier as a filesystem path component in FileStore, allowing out-of-run metadata writes if the service was reachable. The FileStore log_outputs implementation joined caller-supplied model_id into a path without requiring a single safe path component; the fix adds component validation and traversal tests.…
Inspect the ClaimsCase · DiggingBeagle record
MLflow FileStore model_id permitted metadata writes outside the run directory
An MLflow tracking API used an untrusted model identifier as a filesystem path component in FileStore, allowing out-of-run metadata writes if the service was reachable.
MLflow FileStore configurations before the reported 3.18.0 correction, with permission to call log_outputs and filesystem access under the service process.
Assessment profile
Assessment method- Setting
- research demonstration
- Exploitation
- demonstrated
- Evidence
- Grade B
- Remediation
- patch available
- AI role
- AGAINST AI
Basis and provenance
The process could create directories and write meta.yaml outside the intended run subtree; MLflow PR 27131 adds component validation, without evidence of production exploitation. Classified as research_demonstration with exploit status demonstrated. Evidence grade B reflects the original paper or maintainer advisory and its stated limitations, not an independent audit of all deployments. Comparable realized harm is not established, so impact stays unrated; remediation status patch_available is limited to the cited experiment or affected software version. New source-specific boundary: Patch regression results test the FileStore path boundary; they cannot establish whether any previously writable outside-run file was altered in a real installation.
Assessed Nov 15, 2026 using diggingbeagle.assessment/1.
Evidence basis: CLM-R245-MLFLOW-MODELID-FILESTORE-01 · CLM-R245-MLFLOW-MODELID-FILESTORE-02 · CLM-R246-MLFLOW-MODELID-FILESTORE-BOUNDARY · CLM-ENR250-06 · MLflow FileStore model_id permitted metadata writes outside the run directory (original advisory or patch)
At a glance
The patch confirms the validation boundary, while affected release ranges come from a separately reported advisory. Database-backed tracking and a general MLflow server are not automatically vulnerable. No evidence here establishes the number of production victims or a broader failure rate outside the cited evaluation.
Limits and uncertaintyFull account
An MLflow tracking API used an untrusted model identifier as a filesystem path component in FileStore, allowing out-of-run metadata writes if the service was reachable.
The FileStore log_outputs implementation joined caller-supplied model_id into a path without requiring a single safe path component; the fix adds component validation and traversal tests. PR 27131 merged in October and rejects absolute and parent-relative model identifiers in the FileStore path; no downstream model-loading exploit or actual affected customer has been independently demonstrated.
The patch confirms the validation boundary, while affected release ranges come from a separately reported advisory. Database-backed tracking and a general MLflow server are not automatically vulnerable.
The trust boundary is filesystem pathname construction, not model inference. FileStore builds a run output directory using caller-supplied model_id and writes tracking metadata beneath that path. With an absolute or parent-relative component, a caller can steer directory creation and meta.yaml writes outside the intended run subtree, subject to the MLflow process account. The finding is distinct from arbitrary YAML contents or privilege escalation, and later database-backed tracking configurations are not affected through this specific FileStore path.
The merged upstream change adds a single-component guard to log_outputs and tests common normalization cases. The patch supports saying that this route was corrected; it does not prove malicious writes occurred or establish that other FileStore endpoints share this exact fix. Authorized operators should inspect model-output write permissions and installed release, while avoiding extrapolation from this API to every MLflow deployment.
Mechanism and trust boundary
- 01
Entry or starting condition
An attacker can reach an MLflow FileStore tracking API that accepts a model_id for a run output.
Boundary: attacker or research precondition
- 02
Crossing the relevant boundary
The backend previously joined the unvalidated model_id into a filesystem path, allowing parent-relative or absolute components to escape the output directory.
Boundary: mechanism and authorization boundary
- 03
Observed or reported result
The process could create directories and write meta.yaml outside the intended run subtree; MLflow PR 27131 adds component validation, without evidence of production exploitation.
Boundary: measured outcome and evidentiary limit
Claims & evidence
CLM-ENR250-06The MLflow pull request includes regression checks for model identifier path handling. Its reported pre-fix run failed six targeted checks and the patched suite passed 31 checks, supplying maintainer-side validation of FileStore confinement rather than independent evidence about deployed clients or exploitation.supported
Basis: reported finding
- supportsMLflow FileStore model_id permitted metadata writes outside the run directory (original advisory or patch)patch commit artifact
PR #27131 test description and regression output for FileStore model_id validation, merged October 2026
CLM-R245-MLFLOW-MODELID-FILESTORE-01PR 27131 merged in October and rejects absolute and parent-relative model identifiers in the FileStore path; no downstream model-loading exploit or actual affected customer has been independently demonstrated.supported
Basis: reported finding
- supportsMLflow FileStore model_id permitted metadata writes outside the run directory (original advisory or patch)patch commit artifact
PR 27131 summary and validation-test changes
CLM-R245-MLFLOW-MODELID-FILESTORE-02The FileStore log_outputs implementation joined caller-supplied model_id into a path without requiring a single safe path component; the fix adds component validation and traversal tests.supported
Basis: reported finding
- supportsMLflow FileStore model_id permitted metadata writes outside the run directory (original advisory or patch)patch commit artifact
Implementation diff, validate model id as a single path component
CLM-R246-MLFLOW-MODELID-FILESTORE-BOUNDARYMLflow PR 27131 added validation that log_outputs model IDs are single path components, rejecting absolute paths and parent-relative or Windows-normalized traversal; the defect is specific to the FileStore backend and does not itself demonstrate arbitrary code execution.supported
Basis: reported finding
- supportsMLflow FileStore model_id permitted metadata writes outside the run directory (original advisory or patch)patch commit artifact
MLflow pull request 27131, file_store.py log_outputs validation and regression tests
Unknowns and contradictions
- The patch confirms the validation boundary, while affected release ranges come from a separately reported advisory. Database-backed tracking and a general MLflow server are not automatically vulnerable.
- No evidence here establishes the number of production victims or a broader failure rate outside the cited evaluation.
Sources and citation
Material revision history
- Nov 15, 2026 · Canonical change recorded · new in release · revision 251