fix(mlflow): restore SQL-backed registry support

[mldangelo-oai]

Summary

• Restore SQLAlchemy and Alembic to the mlflow, all-ci, and all extras so mlflow-skinny can access SQLite and other SQL-backed model registries again.
• Reintroduce only four packages that were already present before fix(deps): audit all CI extras and upgrade vulnerable packages #1808: SQLAlchemy, Alembic, Greenlet, and Mako.
• Upgrade MLflow's sqlparse dependency to 0.6.0, matching fix(deps): bump sqlparse from 0.5.5 to 0.6.0 #1812 and clearing four currently open security advisories.
• Add regression coverage for every affected extra, the patched SQL parser, and an actual SQLite-backed model registry.

Reproduction

mlflow-skinny explicitly excludes SQL storage dependencies. After #1808 removed full MLflow, creating a registered model against sqlite:///registry.db fails with Model Registry features are not supported by the store. The same operation succeeds with the restored dependency set, including in an isolated environment containing mlflow-skinny without full MLflow.

Validation

• Three new extra-contract regressions failed before the fix and pass afterward.
• The new SQL-parser security-floor regression failed against 0.5.5 and passes with 0.6.0.
• End-to-end SQLite registry creation, model registration, artifact download, and clean scan succeeded in an isolated environment with mlflow-skinny and no full MLflow installation.
• Dependency, workflow, release, Docker, MLflow integration, and organized-asset regression suite: 373 passed, 1 skipped.
• Additional MLflow integration and CLI coverage: 205 passed, 1 skipped.
• Full CI dependency audit: No known vulnerabilities found.
• Repository-wide Ruff lint and formatting, changed-file mypy, and Markdown formatting passed.
• Full local mypy reproduces 14 existing errors in six unchanged files. Native uv lock --check on this macOS host attempts to build an unrelated Linux-only TensorRT placeholder; the exact-head GitHub Linux Lock File Consistency check has already passed.
• Existing default-branch Docker image vulnerabilities remain separately tracked by fix(docker): upgrade vulnerable util-linux runtime packages #1813.

[mldangelo-oai]

@codex review

[chatgpt-codex-connector]

Codex Review Summary

This comment shows the latest Codex review activity on this pull request.

Review	Status	Commit	Review trigger
📝 Code Review	✅ Completed 2026-08-29T15:58:11.247405Z	e0bd50f	Manual request
🔒 Security Review	✅ Completed 2026-08-29T15:58:48.435870Z	e0bd50f	Manual request

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review" or "@codex security review".

Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings.

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 1eae9a76a8

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

[github-actions]

Workflow run and artifacts

Performance Benchmarks

Compared 13 shared benchmarks with a regression threshold of 15%.
Status: 0 regressions, 0 improved, 13 stable, 0 new, 0 missing.
Aggregate shared-benchmark median: 4.318s -> 4.329s (+0.2%).

Workload	Benchmark	Target	Size	Files	Baseline	Current	Change	Status
warm-cache-rescan	tests/benchmarks/test_scan_benchmarks.py::test_scan_warm_cached_repository_rescan	release-candidate	547.3 KiB	32	141.52ms	135.59ms	-4.2%	stable
nested-payload-review	tests/benchmarks/test_picklescan_benchmarks.py::test_picklescan_nested_payload_review[nested_raw]	nested_raw	78 B	1	239.3us	246.3us	+2.9%	stable
direct-malicious-upload	tests/benchmarks/test_picklescan_benchmarks.py::test_picklescan_direct_malicious_upload	malicious_reduce	52 B	1	194.0us	189.8us	-2.2%	stable
mixed-model-repository	tests/benchmarks/test_scan_benchmarks.py::test_scan_release_candidate_repository	release-candidate	547.3 KiB	32	637.77ms	650.34ms	+2.0%	stable
padded-multi-stream-upload	tests/benchmarks/test_picklescan_benchmarks.py::test_picklescan_padded_multi_stream_upload	multi_stream_padded	4.1 KiB	1	309.3us	304.4us	-1.6%	stable
chunked-upload-stream	tests/benchmarks/test_picklescan_benchmarks.py::test_picklescan_chunked_upload_stream	chunked_stream	278.2 KiB	1	118.39ms	117.31ms	-0.9%	stable
nested-payload-review	tests/benchmarks/test_picklescan_benchmarks.py::test_picklescan_nested_payload_review[nested_base64]	nested_base64	98 B	1	262.4us	260.1us	-0.9%	stable
duplicate-heavy-registry	tests/benchmarks/test_scan_benchmarks.py::test_scan_duplicate_registry_snapshot	registry-snapshot	915.2 KiB	13	585.06ms	580.75ms	-0.7%	stable
nested-payload-review	tests/benchmarks/test_picklescan_benchmarks.py::test_picklescan_nested_payload_review[nested_hex]	nested_hex	130 B	1	267.2us	268.8us	+0.6%	stable
suspicious-pickle-intake	tests/benchmarks/test_scan_benchmarks.py::test_scan_suspicious_pickle_intake	suspicious-intake	183.8 KiB	4	147.63ms	148.39ms	+0.5%	stable
rejected-basic-auth-candidates	tests/benchmarks/test_scan_benchmarks.py::test_rejected_basic_auth_candidates_scan_linearly	-	371.1 KiB	1	2.465s	2.473s	+0.4%	stable
clean-training-checkpoint	tests/benchmarks/test_picklescan_benchmarks.py::test_picklescan_clean_training_checkpoint	safe_large	278.2 KiB	1	115.21ms	114.91ms	-0.3%	stable
single-checkpoint-preflight	tests/benchmarks/test_scan_benchmarks.py::test_scan_single_checkpoint_before_load	single_checkpoint.pkl	183.0 KiB	1	106.62ms	106.81ms	+0.2%	stable

[chatgpt-codex-connector]

Security review completed. No security issues were found in this pull request.

Reviewed commit: 1eae9a76a8

View security finding report

Only the user who started this review can view the report in Codex.

ℹ️ About Codex security reviews in GitHub

This is an experimental Codex feature. Security reviews are triggered when:

• You comment "@codex security review"
• A regular code review gets triggered (for example, "@codex review" or when a PR is opened), and you’re opted in so security review runs alongside code review

Once complete, Codex will leave suggestions, or a comment if no findings are found.

[mldangelo-oai]

@codex review

Please review exact head e0bd50f.

[mldangelo-oai]

@codex security review

Please review exact head e0bd50f.

[chatgpt-codex-connector]

Codex Review: Didn't find any major issues. 🚀

Reviewed commit: e0bd50f737

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

[chatgpt-codex-connector]

Security review completed. No security issues were found in this pull request.

Reviewed commit: e0bd50f737

View security finding report

Only the user who started this review can view the report in Codex.

ℹ️ About Codex security reviews in GitHub

This is an experimental Codex feature. Security reviews are triggered when:

• You comment "@codex security review"
• A regular code review gets triggered (for example, "@codex review" or when a PR is opened), and you’re opted in so security review runs alongside code review

Once complete, Codex will leave suggestions, or a comment if no findings are found.