gatekeep-ai 1.0.0

AI-powered governance for development teams — policy-as-code with specialized personas

Gatekeep

AI-powered governance that explains the "why" behind best practices.

Gatekeep catches security issues, cost problems, and design anti-patterns before they reach production. Instead of cryptic linter errors, you get conversational explanations from specialized AI personas who help you learn as you build.

pipx install gatekeep

Alternative installation methods:

• pip install gatekeep (standard Python package)
• Other package managers (Homebrew, Nix, Snap) — let us know if you'd like us to add support!

Quick Start

# Install (recommended: use pipx for isolated environment)
pipx install gatekeep

# Set your OpenRouter API key
export OPENROUTER_API_KEY=sk-or-...

# Ask a persona
gatekeep ask sentinel "Is storing passwords in plaintext safe?"
gatekeep ask auditor "What will 3 Lambda functions at 256MB cost monthly?"
gatekeep ask architect "Should I use DynamoDB or RDS for this use case?"

# Run a full team review
gatekeep review "New payment API with Stripe integration"

# Check deployment readiness
gatekeep deploy "API v2.0 release" --env production

The Personas

Persona	Role	What They Do
🧭 Guide	Triage	Routes your question to the right specialist
👁️ Reviewer	Peer Review	Multi-LLM consensus code review
💰 Auditor	Cost Control	Budget enforcement and optimization
🔒 Sentinel	Security	Vulnerability detection and hardening
🎨 Architect	Design	Architecture patterns and best practices
🧪 Tester	Test Gate	Test environment deployment approval
🛡️ Guardian	Prod Gate	Production deployment approval
📊 Observer	Metrics	Routing, observability, optimization

Each persona has a distinct personality, enforces specific governance rules, and uses the optimal LLM model for their domain.

How It Works

Gatekeep uses a two-layer governance model:

Your Question
     ↓
┌─────────────────────────────────┐
│  PERSONAS (AI Specialists)      │
│  Sentinel, Auditor, Architect…  │
└─────────────────────────────────┘
     ↓                ↓
┌──────────┐  ┌──────────────────┐
│ Your     │  │ Regulatory       │
│ Policies │  │ Standards        │
│          │  │                  │
│ security │  │ OWASP Top 10    │
│ cost     │  │ CIS AWS 2.0     │
│ arch     │  │ GDPR / SOC 2    │
└──────────┘  └──────────────────┘

Layer 1 — Governance (your rules): Customizable YAML policies for security, cost control, and architecture that reflect your organization's standards.

Layer 2 — Standards (external regulations): Built-in support for OWASP Top 10, CIS AWS 2.0, GDPR, and SOC 2. Personas automatically enforce applicable standards.

CLI Commands

gatekeep ask <persona> <question>    # Consult a specific persona
gatekeep review <content>            # Parallel team review (Auditor + Sentinel + Architect)
gatekeep deploy <plan> --env <env>   # Deployment gate check
gatekeep route <question>            # Let Guide pick the right persona
gatekeep personas                    # List all personas
gatekeep standards status            # Show installed standards
gatekeep init                        # Initialize Gatekeep in your project

Project Setup

Initialize Gatekeep in any project to customize governance rules:

cd my-project
gatekeep init

This creates:

• governance/ — Your security, cost, and architecture policies (editable YAML)
• personas/ — Persona definitions (add your own or tweak existing ones)
• standards/ — Regulatory standards (OWASP, CIS, GDPR, SOC 2)
• gatekeep.yaml — Project configuration

Customizing Governance

Edit the YAML files in governance/ to match your organization:

# governance/security.yaml
secrets:
  storage:
    aws:
      primary: "AWS Secrets Manager"
  rotation: "90 days minimum"

encryption:
  in_transit: "TLS 1.2+ required"
  at_rest: "Enable for all storage"

# governance/cost-control.yaml
budgets:
  global:
    monthly_limit: 100
    currency: "USD"

Personas automatically pick up your governance rules and enforce them in their responses.

Requirements

• Python 3.11+
• OpenRouter API key (provides access to Claude, GPT-4, and other models)

Cost Transparency

Gatekeep uses your OpenRouter API key — no data leaves your machine, no subscriptions, no vendor lock-in.

Typical costs:

• Individual developer: $1-5/month
• Small team (5 people): $10-25/month
• You control the spend (set OpenRouter limits)

Compare to:

• GitHub Copilot: $10/user/month ($50/month for 5 people)
• Other AI code review tools: $50-200/month

Gatekeep is cheaper because you only pay for what you use, and we're not trying to extract maximum revenue — we're trying to make governance accessible.

Contributing

See CONTRIBUTING.md for guidelines.

Security

See SECURITY.md for reporting vulnerabilities.

Philosophy

Gatekeep exists because quality code shouldn't be a privilege. Whether you're learning to code, building a side project, or running a startup, you deserve access to the same governance standards that enterprise teams use.

We believe:

• Good practices should be taught, not enforced through fear
• Security and compliance shouldn't require a six-figure budget
• Open source tools should stay open (no bait-and-switch)
• Developers learn best when they understand why, not just what

Commercial Use

Gatekeep is MIT licensed — free for everyone, including commercial use.

If you're building a business on Gatekeep or using it at work, we'd love to hear from you! Consider:

• ⭐ Starring the repo (helps others discover it)
• 💬 Sharing your story (how Gatekeep helped you)
• 💰 Sponsoring development (keeps the project sustainable)
• 🤝 Contributing improvements back (makes it better for everyone)

We're not here to gatekeep governance — we're here to make it accessible.

Support the Project

Gatekeep is built and maintained by developers who care about making quality code accessible. If it's helped you or your team:

• GitHub Sponsors
• Buy Me a Coffee

Every contribution helps us spend more time on Gatekeep and less time on client work.

License

MIT — see LICENSE.

---

Ship safely, not slowly. 🚀