fix(core): group HF cache shard symlinks

[mldangelo-oai]

Summary

• keep validated HuggingFace cache snapshot symlink paths for directory scan routing so sharded filenames remain visible
• still validate shard siblings against resolved HF blob targets before including them in a family
• add regression coverage for HF snapshot shard symlinks being scanned once with a blob-backed shard-family fingerprint

Validation

• env CARGO_HOME=/tmp/modelaudit-cargo-home UV_PROJECT_ENVIRONMENT=/Users/mdangelo/code/modelaudit/.venv UV_CACHE_DIR=/tmp/modelaudit-uv-cache PROMPTFOO_DISABLE_TELEMETRY=1 uv run --no-sync pytest tests/test_core.py::test_directory_scan_groups_hf_cache_sharded_symlinks -q
• env CARGO_HOME=/tmp/modelaudit-cargo-home UV_PROJECT_ENVIRONMENT=/Users/mdangelo/code/modelaudit/.venv UV_CACHE_DIR=/tmp/modelaudit-uv-cache PROMPTFOO_DISABLE_TELEMETRY=1 uv run --no-sync pytest tests/test_core.py -q
• env CARGO_HOME=/tmp/modelaudit-cargo-home UV_PROJECT_ENVIRONMENT=/Users/mdangelo/code/modelaudit/.venv UV_CACHE_DIR=/tmp/modelaudit-uv-cache PROMPTFOO_DISABLE_TELEMETRY=1 uv run --no-sync pytest tests/test_huggingface_symlinks.py -q
• env UV_PROJECT_ENVIRONMENT=/Users/mdangelo/code/modelaudit/.venv UV_CACHE_DIR=/tmp/modelaudit-uv-cache uv run --no-sync ruff format modelaudit/ packages/modelaudit-picklescan/src packages/modelaudit-picklescan/tests tests/
• env UV_PROJECT_ENVIRONMENT=/Users/mdangelo/code/modelaudit/.venv UV_CACHE_DIR=/tmp/modelaudit-uv-cache uv run --no-sync ruff check --fix modelaudit/ packages/modelaudit-picklescan/src packages/modelaudit-picklescan/tests tests/
• env UV_PROJECT_ENVIRONMENT=/Users/mdangelo/code/modelaudit/.venv UV_CACHE_DIR=/tmp/modelaudit-uv-cache uv run --no-sync mypy modelaudit/ packages/modelaudit-picklescan/src packages/modelaudit-picklescan/tests tests/
• env CARGO_HOME=/tmp/modelaudit-cargo-home UV_PROJECT_ENVIRONMENT=/Users/mdangelo/code/modelaudit/.venv UV_CACHE_DIR=/tmp/modelaudit-uv-cache PROMPTFOO_DISABLE_TELEMETRY=1 uv run --no-sync pytest -n auto -m "not slow and not integration" --maxfail=1
• env UV_PROJECT_ENVIRONMENT=/Users/mdangelo/code/modelaudit/.venv UV_CACHE_DIR=/tmp/modelaudit-uv-cache uv run --no-sync ruff check modelaudit/ packages/modelaudit-picklescan/src packages/modelaudit-picklescan/tests tests/
• env UV_PROJECT_ENVIRONMENT=/Users/mdangelo/code/modelaudit/.venv UV_CACHE_DIR=/tmp/modelaudit-uv-cache uv run --no-sync ruff format --check modelaudit/ packages/modelaudit-picklescan/src packages/modelaudit-picklescan/tests tests/

[github-actions]

Workflow run and artifacts

Performance Benchmarks

Compared 12 shared benchmarks with a regression threshold of 15%.
Status: 0 regressions, 0 improved, 12 stable, 0 new, 0 missing.
Aggregate shared-benchmark median: 644.11ms -> 633.66ms (-1.6%).

Workload	Benchmark	Target	Size	Files	Baseline	Current	Change	Status
nested-payload-review	tests/benchmarks/test_picklescan_benchmarks.py::test_picklescan_nested_payload_review[nested_base64]	nested_base64	98 B	1	101.6us	110.0us	+8.2%	stable
nested-payload-review	tests/benchmarks/test_picklescan_benchmarks.py::test_picklescan_nested_payload_review[nested_hex]	nested_hex	130 B	1	105.2us	113.5us	+7.9%	stable
clean-training-checkpoint	tests/benchmarks/test_picklescan_benchmarks.py::test_picklescan_clean_training_checkpoint	safe_large	278.2 KiB	1	14.71ms	15.76ms	+7.1%	stable
chunked-upload-stream	tests/benchmarks/test_picklescan_benchmarks.py::test_picklescan_chunked_upload_stream	chunked_stream	278.2 KiB	1	18.22ms	19.25ms	+5.6%	stable
mixed-model-repository	tests/benchmarks/test_scan_benchmarks.py::test_scan_release_candidate_repository	release-candidate	547.3 KiB	32	264.38ms	254.97ms	-3.6%	stable
single-checkpoint-preflight	tests/benchmarks/test_scan_benchmarks.py::test_scan_single_checkpoint_before_load	single_checkpoint.pkl	183.0 KiB	1	38.34ms	37.55ms	-2.1%	stable
duplicate-heavy-registry	tests/benchmarks/test_scan_benchmarks.py::test_scan_duplicate_registry_snapshot	registry-snapshot	915.2 KiB	13	197.83ms	195.56ms	-1.2%	stable
warm-cache-rescan	tests/benchmarks/test_scan_benchmarks.py::test_scan_warm_cached_repository_rescan	release-candidate	547.3 KiB	32	33.66ms	34.03ms	+1.1%	stable
nested-payload-review	tests/benchmarks/test_picklescan_benchmarks.py::test_picklescan_nested_payload_review[nested_raw]	nested_raw	78 B	1	99.8us	100.4us	+0.6%	stable
suspicious-pickle-intake	tests/benchmarks/test_scan_benchmarks.py::test_scan_suspicious_pickle_intake	suspicious-intake	183.8 KiB	4	75.89ms	75.45ms	-0.6%	stable
padded-multi-stream-upload	tests/benchmarks/test_picklescan_benchmarks.py::test_picklescan_padded_multi_stream_upload	multi_stream_padded	4.1 KiB	1	412.6us	413.6us	+0.2%	stable
direct-malicious-upload	tests/benchmarks/test_picklescan_benchmarks.py::test_picklescan_direct_malicious_upload	malicious_reduce	52 B	1	356.4us	356.5us	+0.0%	stable

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: b7b758079a

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]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 465bf52137

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]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 4f037a7f70

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".