Skip to content

STATUS.md

Latest commit

 

History

History
60 lines (42 loc) · 3.41 KB

STATUS.md

File metadata and controls

60 lines (42 loc) · 3.41 KB

Status

We are claiming 3 badges (Functional, Reusable, and Available) as the following:.

Evaluated - Functional

We believe this artifact satisfies the Functional criteria based on the following evidence:

  • Documented: The artifact is well-documented and includes:

    • System requirements
    • Installation instructions
    • Project structure overview
    • Usage examples
    • Reproduction scripts and step-by-step instructions
    • API documentation generated via Javadoc
    • Developer instructions (develop AccessRefinery in VS Code)

  • Exercisable: The system can be built and executed from source using standard Maven workflows with JDK 17.

  • Complete: The artifact includes all necessary components to reproduce the experimental results reported in the paper:

    • Source code of AccessRefinery
    • Source code of the reimplemented Access Analyzer (used as a baseline, since AWS Access Analyzer is not open source and only exposes a CLI interface)
    • Scripts for invoking AWS Access Analyzer via CLI
    • Three synthetic datasets (real-world datasets are not publicly available due to commercial restrictions)
    • Reproduction and plotting scripts

  • Consistent with the paper: The artifact includes archived experimental results and provides instructions to reproduce all key claims reported in the paper.

Evaluated - Reusable

We believe this artifact satisfies the Reusable criteria for the following reasons:

  • Modular and extensible design: AccessRefinery is designed with clear modular boundaries. In particular, it separates the low-level data structure (MCP) from the higher-level analysis tool. This separation enables users to reuse or extend individual components independently.

  • Reusable library (MCP): MCP is implemented as a standalone Java library. Running mvn package generates a reusable JAR file (mcp-1.0.jar) that can be directly integrated into other projects.

    • It provides a concise and expressive API (e.g., policy.not().and(intent1.or(intent2))).
    • It supports multiple variable types (e.g., regex, prefix, range, and set) through a unified abstraction, making it straightforward to extend to new types.

  • Reusable tool (AccessRefinery): AccessRefinery provides a command-line interface with flexible options for different analysis tasks.

    • Users can easily run experiments or adapt workflows via configurable parameters.
    • The tool supports interchangeable backends (using BDD or SAT solver to represent bitvector constraints), enabling support for additional backends.

  • Reusable baseline implementation (Access Analyzer): We provide a reimplementation of Access Analyzer as a baseline system for evaluation, since AWS Access Analyzer is not open source and only exposes a CLI interface.

    • The reimplementation enables comparison with AccessRefinery under a unified experimental framework.
    • The tool supports interchangeable backends (e.g., Z3 and CVC5 solvers), enabling support for additional backends.

  • Ease of extension: The system is designed to facilitate incremental extensions. For example, adding support for new policy languages only requires extending specific components without modifying the entire system.

  • Documentation and examples: The artifact includes API documentation (via Javadoc) and usage examples..

Available

We believe this artifact satisfies the Available criteria because it is publicly available on Zenodo and GitHub.