FRAI · Framework of Responsible Artificial Intelligence

FRAI (Framework of Responsible Artificial Intelligence) is an open-source toolkit that helps any team — from solo developers to compliance officers — ship AI features responsibly. It walks you through quick questions, scans your code, and generates documentation you can actually hand to stakeholders: implementation checklists, model cards, risk files, evaluation reports, and policy-aware RAG indexes.

Think of FRAI as a safety net for AI launches: it collects the right facts, highlights blind spots, and keeps evidence tidy so production reviews stop feeling like guesswork.

🚀 Getting Started

1. Install FRAI globally:

npm install -g frai

2. Set up your OpenAI API key (required for AI-powered tips):

frai --setup

You only need to do this once per machine or project. Your key is stored securely and never shared.

3. Run FRAI in your project:

frai

Tip: frai is available globally after installing the published npm package. If you're working from a local clone, follow the steps in Local Development to run the CLI from source.

🛠️ CLI Commands & Features

| Command | Description | |------------------------|-------------| | frai | Interactive mode for documenting an AI feature (8-question progressive system) | | frai --scan | Scan codebase for AI/ML code and generate docs | | frai --setup | Set up your OpenAI API key (local/global) | | frai --ci | Run in CI mode (non-interactive) | | frai --help, -h | Show help and usage info | | frai --version, -v | Show current version | | frai --update | Check for new versions of FRAI | | frai --list-docs | List generated documentation files | | frai --clean | Remove generated documentation files | | frai --export-pdf | Export documentation markdown files as PDFs | | frai --show-config | Show API key config status | | frai --key=API_KEY | Provide OpenAI API key directly (one-off use) | | frai --global | Use with --setup to save API key globally | | frai rag index | Index compliance docs into a local vector store | | frai eval | Run baseline evaluation metrics and write reports |

RAG Indexing

frai rag index --input docs/policies --output .frai/compliance-index.json --chunk-size 400

• Scans .md, .txt, .json, .yaml files recursively.
• Generates a lightweight JSON vector store consumable by SDKs and future connectors.

Evaluation Harness

frai eval --outputs runs/outputs.json --references runs/golden.json --report reports/eval --format markdown

• Runs baseline metrics (exact match, toxicity keyword scan, length variance).
• Produces JSON and/or Markdown summaries for CI and governance reviews.

Docs generated:

• checklist.md — Implementation checklist
• model_card.md — Model card
• risk_file.md — Risk & compliance

💡 Features

• Progressive 8-question system: Context-aware, fast, and actionable
• AI-powered recommendations: Get tailored tips for responsible AI
• Comprehensive documentation: Checklist, model card, and risk file
• PDF export: Convert docs to PDF with one command
• Codebase scanning: Detects AI/ML code and generates relevant docs
• Easy setup: One-time API key configuration
• Compliance-aware RAG: Build vector stores from policies for knowledge-grounded guardrails
• Evaluation harness: Run baseline metrics and capture auditable reports

Monorepo Layout

frai/
├─ packages/
│  ├─ frai-cli/      # CLI entry point and command wiring
│  └─ frai-core/     # Reusable services (config, questionnaire, documents, scanners, RAG, eval)
├─ docs/             # Roadmaps, design notes, and feature backlogs
└─ examples/         # Sample AI projects used in tests and demos

🔑 API Key Setup

FRAI requires an OpenAI API key for generating AI-powered tips and documentation. Run:

frai --setup

If you skip this step, FRAI will prompt you to set up your key on first use.

🧑‍💻 Local Development

Run FRAI directly from this repository without publishing:

pnpm install
pnpm --filter frai run build
node packages/frai-cli/dist/index.js --help

Configure an OpenAI key from source

• Interactive CLI:

  node packages/frai-cli/dist/index.js --setup YOUR_KEY

• Manual .env:

  OPENAI_API_KEY=YOUR_KEY

To test the global binary locally (without npm publishing), install the workspace package:

pnpm install --global ./packages/frai-cli
# then:
frai --setup

📖 Learn More

• GitHub Repository
• NPM Package
• AI Feature Backlog
• Evaluation Harness Design

Generated by FRAI - Responsible AI in Minutes