fix(rules): attribute format mismatches to s901

[mldangelo-oai]

Summary

• assign failed core file-format validation checks directly to S901
• prevent incidental terms such as zlib from relabeling those checks as S603
• keep compatible container discrepancies at DEBUG without an S901 code, so severity overrides cannot promote benign wrappers
• retain existing message-based rule inference for legacy scanner checks

Campaign evidence

A SafeTensors/zlib routing collision produced both an S901 mismatch and a duplicate S603 finding. Core emitted Format Validation without a rule code, so ScanResult selected the first catalog regex matching the message. The word zlib matched S603 before the later S901 mismatch rule.

The routing collision is handled separately in #1623. This PR uses a genuine zlib-wrapped malicious payload to isolate rule attribution from that routing fix.

Validation

• focused attribution, malicious-positive, benign-negative, compatible-container, and ONNX routing regressions: 5 passed
• Ruff format and lint pass across required source and test paths
• mypy passes across 473 source files
• full non-slow/non-integration run reached 12,409 passed and 359 skipped; the two local failures were unrelated (read-only HOME cache access and a pre-existing shard-alias test)

[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: 1.409s -> 1.394s (-1.1%).

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	555.6us	534.5us	-3.8%	stable
warm-cache-rescan	tests/benchmarks/test_scan_benchmarks.py::test_scan_warm_cached_repository_rescan	release-candidate	547.3 KiB	32	104.18ms	101.18ms	-2.9%	stable
padded-multi-stream-upload	tests/benchmarks/test_picklescan_benchmarks.py::test_picklescan_padded_multi_stream_upload	multi_stream_padded	4.1 KiB	1	608.3us	599.0us	-1.5%	stable
mixed-model-repository	tests/benchmarks/test_scan_benchmarks.py::test_scan_release_candidate_repository	release-candidate	547.3 KiB	32	486.90ms	479.49ms	-1.5%	stable
chunked-upload-stream	tests/benchmarks/test_picklescan_benchmarks.py::test_picklescan_chunked_upload_stream	chunked_stream	278.2 KiB	1	111.71ms	110.11ms	-1.4%	stable
direct-malicious-upload	tests/benchmarks/test_picklescan_benchmarks.py::test_picklescan_direct_malicious_upload	malicious_reduce	52 B	1	516.7us	509.3us	-1.4%	stable
suspicious-pickle-intake	tests/benchmarks/test_scan_benchmarks.py::test_scan_suspicious_pickle_intake	suspicious-intake	183.8 KiB	4	142.34ms	140.36ms	-1.4%	stable
clean-training-checkpoint	tests/benchmarks/test_picklescan_benchmarks.py::test_picklescan_clean_training_checkpoint	safe_large	278.2 KiB	1	107.36ms	106.81ms	-0.5%	stable
nested-payload-review	tests/benchmarks/test_picklescan_benchmarks.py::test_picklescan_nested_payload_review[nested_raw]	nested_raw	78 B	1	537.8us	535.6us	-0.4%	stable
nested-payload-review	tests/benchmarks/test_picklescan_benchmarks.py::test_picklescan_nested_payload_review[nested_hex]	nested_hex	130 B	1	566.0us	568.3us	+0.4%	stable
single-checkpoint-preflight	tests/benchmarks/test_scan_benchmarks.py::test_scan_single_checkpoint_before_load	single_checkpoint.pkl	183.0 KiB	1	69.99ms	69.71ms	-0.4%	stable
duplicate-heavy-registry	tests/benchmarks/test_scan_benchmarks.py::test_scan_duplicate_registry_snapshot	registry-snapshot	915.2 KiB	13	383.64ms	383.40ms	-0.1%	stable

[mldangelo-oai]

@codex review

Exact current head: 355226d65aa559be809a00bb6b5d317b391c9ec9. This supersedes the mistyped full SHA in my preceding comment.

[chatgpt-codex-connector]

Codex Review: Didn't find any major issues. Keep them coming!

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".

[mldangelo-oai]

@codex review

Exact current head SHA: f0a1998. This head includes the latest origin/main via a normal merge.

[chatgpt-codex-connector]

Codex Review: Didn't find any major issues. Swish!

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".

[mldangelo-oai]

@codex review

Exact current head SHA: ce771f5. This head includes latest origin/main commit e8dc55e via a normal merge.

[mldangelo-oai]

@codex review

Exact current head SHA: 1d13b12. This head includes latest origin/main commit c769097 via a normal merge.

[chatgpt-codex-connector]

Codex Review: Didn't find any major issues. Nice work!

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".

[mldangelo-oai]

@codex review

Exact current head SHA: 2e3ee8c. This head includes latest origin/main commit e407fb4 via a normal merge.

[chatgpt-codex-connector]

Codex Review: Didn't find any major issues. Another round soon, please!

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".