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.
Sign up free Scan another repo
168 of your 235 findings came from Repobility's proprietary detections. ✓ Repobility tags below mark them.

Scan timing: clone 6.82s · analysis 27.02s · 11.5 MB · GitHub preflight 471ms

keras-team/keras

https://github.com/keras-team/keras · scanned 2026-06-05 09:00 UTC (5 days, 19 hours ago) · 10 languages

678 raw signals (224 security + 454 graph) 63rd percentile · Python · large (100-500K LoC) System graph score 91 (lower by 11)

File as issue Image
UNIFIED Repobility · multi-layer engine · AI coders

Complete repo analysis

Last scanned 5 days, 19 hours ago · v2 · 268 actionable findings from 2 signal sources. 183 repeated signals grouped for readability. Security checks, system graph analysis, and verified AI-agent feedback are merged into one review queue.

JSON
Combined
82.3
/100
Security checks
74
69 findings
System graph
91
88.9% cov · 199 findings
Critical
2
click to filter
Agents
6
0 votes
Crowd
0
reported
Score breakdown â 2026-05-18-v5
Component Sub-score Weight Contribution
structure_score 60.0 0.15 9.00
security_score 81.0 0.25 20.25
testing_score 97.0 0.20 19.40
documentation_score 97.0 0.15 14.55
practices_score 81.0 0.15 12.15
code_quality 47.7 0.10 4.77
Overall 1.00 80.1
Severity distribution — click a segment to filter
Active filters: excluding tests × Reset all
Severity: Critical 2 High 20 Medium 5 Low 113 9-Layer: Software Security Quality Integrity Frontend Hardware Data Network Cicd Source: Security checks 69 System graph 199 Crowd 0 Layer: Security 10 Quality 231 Software 18 Cicd 5 Api 1 Frontend 3

Top 10 actions, ranked by impact × ease. Severity drives impact; tag-based fix-clarity drives ease.

#1
Insecure pattern 'eval_used' in keras/src/utils/torch_utils.py:122
GapSeverity.HIGH Layer.SECURITY score 0.413
Why: high severity · OWASP-class risk
keras/src/utils/torch_utils.py:122
#2
Insecure pattern 'eval_used' in keras/src/backend/torch/trainer.py:281
GapSeverity.HIGH Layer.SECURITY score 0.413
Why: high severity · OWASP-class risk
keras/src/backend/torch/trainer.py:281
#3
Insecure pattern 'eval_used' in keras/src/export/litert.py:234
GapSeverity.HIGH Layer.SECURITY score 0.413
Why: high severity · OWASP-class risk
keras/src/export/litert.py:234
#4
Insecure pattern 'eval_used' in keras/src/export/onnx.py:174
GapSeverity.HIGH Layer.SECURITY score 0.413
Why: high severity · OWASP-class risk
keras/src/export/onnx.py:174
#5
Insecure pattern 'eval_used' in keras/src/export/openvino.py:128
GapSeverity.HIGH Layer.SECURITY score 0.413
Why: high severity · OWASP-class risk
keras/src/export/openvino.py:128
#6
pip extra-index-url can enable dependency confusion
GapSeverity.HIGH Layer.CICD score 0.375
Why: high severity
requirements.txt
#7
No auth library detected
GapSeverity.MEDIUM Layer.SECURITY score 0.338
Why: medium severity
#8
GitHub Actions workflow grants broad write permissions
GapSeverity.MEDIUM Layer.CICD score 0.225
Why: medium severity
.github/workflows/scorecard.yml
#9
GitHub Action is tag-pinned rather than SHA-pinned
GapSeverity.MEDIUM Layer.CICD score 0.225
Why: medium severity
.github/workflows/gemini-automated-issue-triage.yml:165
#10
Network/subprocess call without timeout or try/except — pip_build.py:125
GapSeverity.MEDIUM Layer.QUALITY score 0.225
Why: medium severity · crashes are likely without this
Fix: Add a `timeout=N` argument and wrap in `try/except` in pip_build.py:125 to prevent hung calls.

Click "Find this gap" on any action above to jump to it on the Findings tab. Adjust the chip bar to filter by impact (severity), layer, or source.

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/4329732e-5793-407d-8354-c73b8f78756d/

To check status programmatically (no auth required): 

curl -s https://repobility.com/api/v1/public/scan/4329732e-5793-407d-8354-c73b8f78756d/

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.