Tabular Classification
Kernels
Joblib
Scikit-learn
kernel
governance
formulas
proof-status
lean4
lambda-aggregate
surrogate
doi:10.5281/zenodo.19944926
Instructions to use SZLHOLDINGS/szl-formulas with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Kernels
How to use SZLHOLDINGS/szl-formulas with Kernels:
# !pip install kernels from kernels import get_kernel kernel = get_kernel("SZLHOLDINGS/szl-formulas") - Scikit-learn
How to use SZLHOLDINGS/szl-formulas with Scikit-learn:
from huggingface_hub import hf_hub_download import joblib model = joblib.load( hf_hub_download("SZLHOLDINGS/szl-formulas", "sklearn_model.joblib") ) # only load pickle files from sources you trust # read more about it here https://skops.readthedocs.io/en/stable/persistence.html - Notebooks
- Google Colab
- Kaggle
Publish szl-formulas: honest stdlib-only offline replay kernel (14/14 & 13/13 trio)
59baf06 verified Security Policy
Reporting a Vulnerability
Do NOT open a public issue for security vulnerabilities.
Please report security vulnerabilities via email to security@szlholdings.com with:
- Description of the vulnerability
- Steps to reproduce
- Potential impact assessment
- Any suggested mitigations
Response SLA
| Severity | Initial Response | Resolution Target |
|---|---|---|
| Critical | 24 hours | 7 days |
| High | 48 hours | 30 days |
| Medium | 5 business days | 90 days |
| Low | 10 business days | 180 days |
We follow a 90-day responsible disclosure policy.
Scope notes for this artifact
- No trained weights. This repository ships pure-Python, stdlib-only code — there is no
.safetensors/.bin/.pt/.ggufweight artifact to poison or exfiltrate. - Offline by construction. The formulas are pure and no-I/O; the governed-loop composer performs
no network access. Inputs are validated — bad weights, non-probability distributions, or
k > nraiseValueErrorrather than returning a coerced answer. - Honest proof surface.
PROOF_STATUSis mirrored verbatim from the canonical dataset; the locked-proven canonical set is fixed at exactly 8 and never inflated. APROVEN(...)obligation tag is never upgraded into a locked-proven claim. Λ uniqueness stays Conjecture 1 (open).
Supply-Chain Security
- DCO required — all commits carry
Signed-off-by:trailers per Linux Foundation DCO - SLSA Build Level 1 — honest; not L2/L3
Contact
- Security disclosures: security@szlholdings.com
- General: hello@szlholdings.com
- Website: https://szlholdings.com
This policy follows the OpenSSF Vulnerability Disclosure Guide.