tensorflow/models

Public scan — anyone with this URL can view this analysis. Sign up to track your own repos privately, run scheduled re-scans, and get AI fix prompts via your dashboard.

172 of your 294 findings came from Repobility's proprietary detections. ✓ Repobility tags below mark them.

Upstream (GitHub) caused delay on this scan — not Repobility.

GitHub API rate-limited (HTTP 403) — preflight skipped, fell back to direct git clone.
Clone from GitHub took 94.5s for a 106.5 MB repo slow.
Repobility's analysis ran in 41.11s after the clone landed.

https://github.com/tensorflow/models · scanned 2026-06-05 07:36 UTC (1 week, 1 day ago) · 10 languages

1332 raw signals (274 security + 1058 graph) 11/13 scanners ran 59th percentile · Python · huge (>500K LoC) System graph score 61 (higher by 17)

File as issue Image

UNIFIED Repobility · multi-layer engine · AI coders

Complete repo analysis

Last scanned 1 week, 1 day ago · v2 · 612 actionable findings from 2 signal sources. 191 repeated signals grouped for readability. Security checks, system graph analysis, and verified AI-agent feedback are merged into one review queue.

JSON

88.9% cov · 511 findings

Score breakdown â 2026-05-18-v5

Component	Sub-score	Weight	Contribution
`structure_score`	40.0	0.15	6.00
`security_score`	100.0	0.25	25.00
`testing_score`	82.0	0.20	16.40
`documentation_score`	100.0	0.15	15.00
`practices_score`	65.0	0.15	9.75
`code_quality`	56.0	0.10	5.60
Overall		1.00	77.8

security_score may be inflated — optional security scanners were skipped on this fast scan

Severity distribution — click a segment to filter

Active filters: excluding tests × Reset all

Severity: Critical 3 High 48 Medium 28 Low 346 9-Layer: Software Security Quality Integrity Frontend Hardware Data Network Cicd Source: Security checks 101 System graph 511 Crowd 0 Layer: Software 77 Quality 460 Cicd 8 Security 56 Frontend 2 Hardware 8 Api 1

Exclude dismissed / FP Exclude test files

Bug-class explainers. Each card groups findings of the same shape — these are the patterns most likely to ship to prod and reappear in future scans unless you systematically fix the cause, not just the instance.

Duplicates & near-duplicates 4 findings

What it is: Same function copy-pasted into multiple modules with minor variations.

Why it matters: Each copy drifts independently — bug fixes apply to one, miss the others.

How AI causes it: AI completes the same pattern in each file rather than refactoring to a shared helper.

Fix approach: Extract the duplicated logic into the most general module both call sites already import. Add tests at the helper level.

4 matching findings on this repo

low Near-duplicate function bodies in 2 places repo-level
low Near-duplicate function bodies in 4 places
low Near-duplicate function bodies in 5 places
low Near-duplicate function bodies in 3 places

View all duplicates & near-duplicates findings →

Legacy markers 24 findings

What it is: TODO, FIXME, XXX, HACK comments. Often indicate a known-broken path the author meant to fix.

Why it matters: Each marker is an unfinished thought. Production code shouldn't ship with debt that's documented but not tracked.

How AI causes it: AI mirrors the style of the codebase, so existing TODOs propagate into new code.

Fix approach: Convert each into a ticket. Delete the comment when the ticket lands. Use a pre-commit hook to block new TODOs without an issue link.

12 matching findings on this repo

high Insecure pattern 'eval_used' in official/legacy/transformer/transformer_main_te… official/legacy/transformer/transformer_main_test…:189
high Insecure pattern 'eval_used' in official/legacy/transformer/transformer_main.py… official/legacy/transformer/transformer_main.py:352
low Stub function `restore_from_objects` (body is just `pass`/`return`) — research/…
info Commented-code block (5 lines) in official/legacy/xlnet/preprocess_classificati…
low Old/deprecated-named symbol `transformer_v2` in official/legacy/transformer/tra…
info Commented-code block (7 lines) in official/legacy/transformer/data_download.py:…
medium Network/subprocess call without timeout or try/except — official/legacy/transfo…
low Old/deprecated-named symbol `transformer_v2` in official/legacy/transformer/tra…
low Old/deprecated-named symbol `transformer_v2` in official/legacy/transformer/tra…
low Old/deprecated-named symbol `softmax_cross_entropy_with_logits_v2` in official/…
info Commented-code block (6 lines) in official/legacy/bert/run_pretraining.py:165
info Commented-code block (6 lines) in official/legacy/bert/run_squad_helper.py:258

View all legacy markers findings →

Commented-out code 117 findings

What it is: Lines of source that were intentionally disabled but never deleted.

Why it matters: Git already remembers history — commented code rots, becomes wrong, and adds noise to diffs.

How AI causes it: AI sometimes comments out broken code instead of fixing it. Reviewers approve out of inertia.

Fix approach: Delete. Trust `git log`. If you really need to remember, save it in a notes file under `docs/`.

12 matching findings on this repo

info Commented-code block (8 lines) in research/slim/datasets/build_imagenet_data.py…
info Commented-code block (7 lines) in research/slim/nets/inception_v2.py:112
info Commented-code block (6 lines) in research/slim/nets/resnet_utils.py:269
info Commented-code block (6 lines) in research/slim/nets/mobilenet/mobilenet.py:224
info Commented-code block (7 lines) in research/attention_ocr/python/model.py:665
info Commented-code block (8 lines) in research/attention_ocr/python/data_provider.p…
info Commented-code block (5 lines) in research/vid2depth/model.py:186
info Commented-code block (5 lines) in research/vid2depth/ops/icp_train_demo.py:42
info Commented-code block (5 lines) in research/vid2depth/dataset/dataset_loader.py:…
info Commented-code block (5 lines) in research/audioset/yamnet/features.py:129
info Commented-code block (5 lines) in research/audioset/vggish/vggish_smoke_test.py…
info Commented-code block (6 lines) in research/audioset/vggish/vggish_slim.py:64

View all commented-out code findings →

Config drift 19 findings

What it is: Settings duplicated across env files, Docker compose, K8s, and code defaults, all with slightly different values.

Why it matters: Production behaviour depends on whichever copy your loader reads first. Subtle bugs in staging that don't reproduce in dev.

How AI causes it: AI writes new config from memory rather than reading the existing source.

Fix approach: Pick one source of truth (env vars + a settings module). Have every other place import from there. Lint for duplicates in CI.

12 matching findings on this repo

low File has no detected symbols: research/rebar/config.py
low File has no detected symbols: official/nlp/configs/experiment_configs.py
low File has no detected symbols: official/projects/waste_identification_ml/circula…
low Very large file: research/object_detection/utils/config_util.py (1271 lines)
low Very large file: research/object_detection/utils/config_util_test.py (1084 line…
low Old/deprecated-named symbol `bert_v2` in official/nlp/configs/encoders.py:330
info Commented-code block (5 lines) in official/projects/videoglue/configs/spatiotem…
info Commented-code block (8 lines) in official/projects/maskconver/configs/multisca…
info Commented-code block (8 lines) in official/projects/maskconver/configs/maskconv…
info Commented-code block (7 lines) in official/projects/panoptic/configs/panoptic_m…
low Old/deprecated-named symbol `resnet50_v1` in official/projects/panoptic/configs…
low Stub function `set_quantize_weights` (body is just `pass`/`return`) — official/…

View all config drift findings →

For AI agents: Voting guide (TP/FP) MCP manifest Stdio wrapper SARIF Integrate Findings queue Vote TP/FP on findings to calibrate the engine.

For AI agents + API integrations

Email me when this repo regresses

Free. We re-scan periodically; new criticals → your inbox. No signup required for the scan itself.

API access

This page is publicly accessible at: https://repobility.com/scan/e050a7b9-f185-48e3-8563-c0243fa62a6b/

To check status programmatically (no auth required):

curl -s https://repobility.com/api/v1/public/scan/e050a7b9-f185-48e3-8563-c0243fa62a6b/

Important — please don't re-submit the same URL repeatedly. The submission endpoint is idempotent: re-submitting the same git URL returns this same scan_token, not a new one. To re-scan this repo, sign up free and use the dashboard.