AI Harness Scorecard
Version updated for https://github.com/markmishaev76/ai-harness-scorecard to version v1.1.0.
- This action is used across all versions by 3 repositories.
Action Type
This is a Composite action.
Go to the GitHub Marketplace to find the latest changes.
Action Summary
The AI Harness Scorecard GitHub Action assesses a repository’s engineering safeguards to ensure reliable and safe AI-assisted development. It automates the evaluation of key areas such as architectural documentation, mechanical constraints, testing practices, code review enforcement, and AI-specific safeguards, providing a comprehensive grade and actionable insights. This tool helps teams identify gaps, improve their development processes, and ensure that AI-generated code adheres to high-quality standards.
Release notes
AI Harness Scorecard v1.1.0
New Features
- Java ecosystem support: linting detection, formatting via Spotless and Checkstyle, dependency auditing, and property-based testing checks
- Java fuzz testing: detect Jazzer/jazzer-junit in Gradle projects
- Kotlin support: Gradle Spotless detection now covers Kotlin projects
- Package manager script resolution: CI linter check now resolves
npm run,yarn,pnpm, andbunscripts to their underlying commands - Resources knowledge base: curated collection of papers, books, talks, and specs behind the scorecard’s methodology
Bug Fixes
- Formatter enforcement check no longer incorrectly treats
checkstyle.xmlas a formatter - JSON script loading is more robust with better error handling
- Scorecard workflow uses
create-pull-requestfor reliable updates
Maintenance
- Pre-commit hooks added (ruff, mypy, custom hooks on push)
- GitHub Actions bumped:
actions/checkoutv6,actions/setup-pythonv6,actions/upload-artifactv7,peter-evans/create-pull-requestv8 - Type-hint-only imports moved into
TYPE_CHECKINGblocks - Check development guide added to CONTRIBUTING.md
Usage
- uses: markmishaev76/ai-harness-scorecard@v1
id: scorecard
Users on @v1 automatically get this update.