pretty_name: SARMF-Bench
license: cc-by-4.0
tags:
- smart-contracts
- blockchain-security
- static-analysis
- benchmark
- reproducible-research
- cybersecurity
- ethereum
- software-security
language:
- en
SARMF-Bench
SARMF-Bench is a reproducible smart contract vulnerability benchmark designed for controlled evaluation of security analysis tools.
Overview
The benchmark consists of five intentionally engineered Solidity contracts aligned with canonical SWC vulnerability classes:
- SC01 – Reentrancy (SWC-107)
- SC02 – Integer Overflow Behavior (SWC-101)
- SC03 – Access Control Weakness (SWC-105)
- SC04 – Unchecked External Call (SWC-104)
- SC05 – Denial-of-Service via Unbounded Loop (SWC-113)
Each contract is paired with machine-readable static analysis outputs generated using Slither v0.11.5 in a version-locked Linux environment.
Primary Repository
GitHub: https://github.com/profmohit-edu/sarmf-framework
DOI-Backed Archives
Zenodo Software DOI: https://doi.org/10.5281/zenodo.18754015
Harvard Dataverse Dataset DOI: https://doi.org/10.7910/DVN/0SP3OO
Mendeley Data Mirror: https://doi.org/10.17632/d3zkh9cmpn.1
OSF Project: https://doi.org/10.17605/OSF.IO/EJWDC
Reproducibility Protocol: https://doi.org/10.17504/protocols.io.bp216eyxdgqe/v1
Reproducibility Design
- Deterministic contract design
- Explicit SWC taxonomy alignment
- Version-locked toolchain (Solidity 0.8.x, Slither v0.11.5)
- Machine-readable JSON outputs
- Cross-archived infrastructure (Zenodo + HAL + Software Heritage)
Intended Use
SARMF-Bench supports benchmarking of:
- Static analyzers
- Symbolic execution engines
- Fuzzers
- AI-assisted smart contract security systems
- Security education experiments
Citation
If you use SARMF-Bench, please cite the Zenodo DOI and related archival identifiers listed above.