Skip to content

Fix arbitrary file read via prompt tag validation bypass in Model Registry#20833

Merged
TomeHirata merged 4 commits intomlflow:masterfrom
TomeHirata:fix/prompt-source-validation
Feb 17, 2026
Merged

Fix arbitrary file read via prompt tag validation bypass in Model Registry#20833
TomeHirata merged 4 commits intomlflow:masterfrom
TomeHirata:fix/prompt-source-validation

Conversation

@TomeHirata
Copy link
Collaborator

@TomeHirata TomeHirata commented Feb 16, 2026

Related Issues/PRs

Fixes #20818

What changes are proposed in this pull request?

See this ticket for the security vulnerability

How is this PR tested?

  • Existing unit/integration tests
  • New unit/integration tests
  • Manual tests

Does this PR require documentation update?

  • No. You can skip the rest of this section.
  • Yes. I've updated:
    • Examples
    • API references
    • Instructions

Does this PR require updating the MLflow Skills repository?

  • No. You can skip the rest of this section.
  • Yes. Please link the corresponding PR or explain how you plan to update it.

Release Notes

Is this a user-facing change?

  • No. You can skip the rest of this section.
  • Yes. Give a description of this change to be included in the release notes for MLflow users.

What component(s), interfaces, languages, and integrations does this PR affect?

Components

  • area/tracking: Tracking Service, tracking client APIs, autologging
  • area/models: MLmodel format, model serialization/deserialization, flavors
  • area/model-registry: Model Registry service, APIs, and the fluent client calls for Model Registry
  • area/scoring: MLflow Model server, model deployment tools, Spark UDFs
  • area/evaluation: MLflow model evaluation features, evaluation metrics, and evaluation workflows
  • area/gateway: MLflow AI Gateway client APIs, server, and third-party integrations
  • area/prompts: MLflow prompt engineering features, prompt templates, and prompt management
  • area/tracing: MLflow Tracing features, tracing APIs, and LLM tracing functionality
  • area/projects: MLproject format, project running backends
  • area/uiux: Front-end, user experience, plotting, JavaScript, JavaScript dev server
  • area/build: Build and test infrastructure for MLflow
  • area/docs: MLflow documentation pages

How should the PR be classified in the release notes? Choose one:

  • rn/none - No description will be included. The PR will be mentioned only by the PR number in the "Small Bugfixes and Documentation Updates" section
  • rn/breaking-change - The PR will be mentioned in the "Breaking Changes" section
  • rn/feature - A new user-facing feature worth mentioning in the release notes
  • rn/bug-fix - A user-facing bug fix worth mentioning in the release notes
  • rn/documentation - A user-facing documentation change worth mentioning in the release notes

Should this PR be included in the next patch release?

Yes should be selected for bug fixes, documentation updates, and other small changes. No should be selected for new features and larger changes. If you're unsure about the release classification of this PR, leave this unchecked to let the maintainers decide.

What is a minor/patch release?
  • Minor release: a release that increments the second part of the version number (e.g., 1.2.0 -> 1.3.0).
    Bug fixes, doc updates and new features usually go into minor releases.
  • Patch release: a release that increments the third part of the version number (e.g., 1.2.0 -> 1.2.1).
    Bug fixes and doc updates usually go into patch releases.
  • Yes (this PR will be cherry-picked and included in the next patch release)
  • No (this PR will be included in the next minor release)

…istry

Reject local filesystem sources (file:// URIs and absolute paths) when
creating model versions with the prompt tag, preventing attackers from
using the prompt code path to bypass source validation and read arbitrary
server files via the artifact download endpoint.

Co-Authored-By: Claude <noreply@anthropic.com>
Signed-off-by: Tomu Hirata <tomu.hirata@gmail.com>
Copilot AI review requested due to automatic review settings February 16, 2026 05:02
@TomeHirata TomeHirata added the team-review Trigger a team review request label Feb 16, 2026
@github-actions github-actions bot added area/prompts MLflow Prompt Registry and Optimization rn/bug-fix Mention under Bug Fixes in Changelogs. v3.10.0 size/S Small PR (10-49 LoC) labels Feb 16, 2026
@github-actions
Copy link
Contributor

🛠 DevTools 🛠

Install mlflow from this PR

# mlflow
pip install git+https://github.com/mlflow/mlflow.git@refs/pull/20833/merge
# mlflow-skinny
pip install git+https://github.com/mlflow/mlflow.git@refs/pull/20833/merge#subdirectory=libs/skinny

For Databricks, use the following command:

%sh curl -LsSf https://raw.githubusercontent.com/mlflow/mlflow/HEAD/dev/install-skinny.sh | sh -s pull/20833/merge

Copy link
Contributor

Copilot AI left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Pull request overview

This PR addresses a security issue in MLflow Model Registry where creating a prompt model version could bypass source validation, potentially enabling arbitrary local file reads via crafted source values.

Changes:

  • Add prompt-specific validation in CreateModelVersion to reject local filesystem sources and validate against relative-path traversal.
  • Add a unit test ensuring prompt model version creation rejects local source paths.

Reviewed changes

Copilot reviewed 2 out of 2 changed files in this pull request and generated 2 comments.

File Description
mlflow/server/handlers.py Enforces non-local prompt sources and applies traversal validation for prompt model version creation.
tests/server/test_handlers.py Adds coverage to ensure prompt model version creation rejects local filesystem source inputs.

@github-actions
Copy link
Contributor

github-actions bot commented Feb 16, 2026

Documentation preview for f98f125 is available at:

More info
  • Ignore this comment if this PR does not change the documentation.
  • The preview is updated when a new commit is pushed to this PR.
  • This comment was created by this workflow run.
  • The documentation was built by this workflow run.

TomeHirata and others added 2 commits February 16, 2026 14:30
- Wrap is_local_uri() in try/except to normalize errors to
  INVALID_PARAMETER_VALUE (400) instead of INTERNAL_ERROR (500)
  for invalid prompt sources like file://remote-host/path
- Add test cases for file://remote-host bypass attempts and
  encoded path traversal (..%2f) in non-local URIs

🤖 Generated with Claude Code

Co-Authored-By: Claude <noreply@anthropic.com>
Signed-off-by: Tomu Hirata <tomu.hirata@gmail.com>
The previous validation using is_local_uri() rejected schemeless sources
like "prompt-template" and "dummy-source" that are used internally as
placeholders for prompt model versions. Narrowed the check to only block
file:// URIs and absolute paths, and only run traversal validation on
sources with an actual URL scheme.

Co-Authored-By: Claude <noreply@anthropic.com>
Signed-off-by: Tomu Hirata <tomu.hirata@gmail.com>
parsed = urllib.parse.urlparse(source)
if parsed.scheme == "file" or (parsed.scheme == "" and source.startswith("/")):
raise MlflowException(
f"Invalid model version source: '{source}'. "
Copy link
Member

@harupy harupy Feb 17, 2026

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Suggested change
f"Invalid model version source: '{source}'. "
f"Invalid prompt source: '{source}'. "

can we replace model version with prompt?

)
resp = _create_model_version()
assert resp.status_code == 400
data = json.loads(resp.get_data())
Copy link
Member

@harupy harupy Feb 17, 2026

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

resp doesn't have a method to get dict?

)
resp = _create_model_version()
assert resp.status_code == 400
data = json.loads(resp.get_data())
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Copy link
Member

@harupy harupy left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Left comments. LGTM once they are addressed!

- Use "Invalid prompt source" instead of "Invalid model version source"
  in the prompt-specific error message
- Use resp.get_json() instead of json.loads(resp.get_data()) in tests

🤖 Generated with Claude Code

Co-Authored-By: Claude <noreply@anthropic.com>
Signed-off-by: Tomu Hirata <tomu.hirata@gmail.com>
@TomeHirata TomeHirata enabled auto-merge February 17, 2026 06:02
@TomeHirata TomeHirata added this pull request to the merge queue Feb 17, 2026
Merged via the queue into mlflow:master with commit 6e801f4 Feb 17, 2026
49 checks passed
@TomeHirata TomeHirata deleted the fix/prompt-source-validation branch February 17, 2026 06:42
@github-actions github-actions bot added size/M and removed size/S Small PR (10-49 LoC) labels Feb 17, 2026
daniellok-db pushed a commit to daniellok-db/mlflow that referenced this pull request Feb 20, 2026
…istry (mlflow#20833)

Signed-off-by: Tomu Hirata <tomu.hirata@gmail.com>
Co-authored-by: Claude <noreply@anthropic.com>
daniellok-db pushed a commit that referenced this pull request Feb 20, 2026
…istry (#20833)

Signed-off-by: Tomu Hirata <tomu.hirata@gmail.com>
Co-authored-by: Claude <noreply@anthropic.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

area/prompts MLflow Prompt Registry and Optimization rn/bug-fix Mention under Bug Fixes in Changelogs. size/M team-review Trigger a team review request v3.10.0

Projects

None yet

Development

Successfully merging this pull request may close these issues.

[BUG] Security Vulnerability

3 participants