KurniaKadir/Web3-Dataset

Web3Coders Smart Contracts Dataset – 2026 Edition

Version 2.0 (August 2026) – The largest curated collection of production-grade smart contracts with integrated security audits, gas profiles, vulnerability annotations, and Halal compliance tags for Shariah‑aware Web3 development.

Dataset Description

A comprehensive, multi‑chain dataset of smart contracts (Solidity, Vyper, Rust, and Move) with:

• —Full source code, bytecode, and ABI
• —Security audit outcomes (Slither, Mythril, and manual review)
• —Gas consumption profiles (execution cost, deployment cost)
• —Halal compliance annotations based on AAOIFI Shariah standards, including Riba, Gharar, and Maisir detection
• —Fatwa references and compliance reports for Islamic finance applications

This dataset powers the HalalChain Web3 Coders AI assistant, enabling Shariah‑compliant code generation and auditing.

• —Homepage: https://huggingface.co/datasets/KurniaKadir/Web3Coders
• —Repository: https://github.com/kk666679/turbo-doodle

Intended Uses

✔️ Smart Contract Development – Train and fine‑tune code generation models ✔️ Security Analysis – Build vulnerability detection and audit automation ✔️ Gas Optimization Research – Benchmark and optimise gas usage ✔️ ABI & API Generation – Synthesise interfaces and client libraries ✔️ Shariah‑Compliant DeFi – Identify and enforce Islamic finance rules in DeFi protocols

Dataset Structure

.
├── data/                    # Processed dataset (JSONL format)
│   ├── train.parquet
│   ├── validation.parquet
│   └── test.parquet
├── source_contracts/        # Original Solidity/Vyper/Move/Rust files
├── audits/                  # PDF/JSON security reports
├── gas_profiles/            # Execution cost analyses
├── halal_compliance/        # AAOIFI compliance status & fatwa references
└── metadata/                # Contract labels, categories, and tags

Usage Example

from datasets import load_dataset

# Load the latest version (August 2026)
dataset = load_dataset(
    "Web3Coders/smart-contracts-dataset",
    revision="v2.0.0",
    split="train[:20%]+validation"
)

# Filter for Halal‑compliant contracts
halal_contracts = dataset.filter(
    lambda x: x["halal_compliance"] == "compliant"
)

# Access security‑audited contracts
secure_contracts = dataset.filter(
    lambda x: x["security_audit"] == "pass"
)

# Get gas estimates for optimisation
gas_data = dataset.select_columns(["source_code", "gas_estimates"])

Dataset Creation (2026 Update)

Sources (Expanded)

Total: 84,217 unique smart contracts (↑ 97% from 2024)

Annotation Process

1. 1.Security Analysis:

   slither.detect_all(solc_version="0.8.26")
   mythril.analyze_contract(bytecode, timeout=120)

1. 1.Gas Profiling:

   hardhat test --gas --parallel

1. 1.Halal Compliance Checking (new):

   from halalchain import ComplianceChecker
   checker = ComplianceChecker()
   status, report = checker.check_contract(source_code, rules=["no-riba", "no-gharar"])

1. 1.Fatwa Integration:
2. 2.AAOIFI Standards (2024 update)
3. 3.Fatwa references from HalalChain knowledge base
4. 4.Manual review by Shariah advisors

Ethical Considerations

• —Privacy: All sensitive data (private keys, addresses) removed.
• —Licensing: Clear provenance tracking; only permissively licensed contracts included.
• —Bias Mitigation: Balanced representation across DeFi, NFTs, DAOs, and Islamic finance.
• —Transparency: Compliance annotations are reviewed by a multi‑disciplinary team.

Citation

If you use this dataset, please cite:

@misc{web3coders2026dataset,
  title={Web3Coders Smart Contracts Dataset – 2026 Edition},
  author={Web3Coders Team and HalalChain Contributors},
  year={2026},
  month={August},
  publisher={Hugging Face},
  url={https://huggingface.co/datasets/Web3Coders}
}

License

Apache 2.0 – See LICENSE for full terms.

Part of the HalalChain Web3 Coders ecosystem – Empowering Shariah‑first decentralized innovation.