Skip to content

LLM Software Development Standards Start any project right in 30 seconds. Battle-tested standards from real production systems.

License

Notifications You must be signed in to change notification settings

williamzujkowski/standards

Folders and files

NameName
Last commit message
Last commit date

Latest commit

Β 

History

176 Commits
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 

Repository files navigation

πŸš€ Software Development Standards

Anthropic Skills

Start any project right in 30 seconds. Based on industry best practices and NIST guidelines.

Why This Repository?

You're starting a new project. You need to make dozens of decisions: project structure, testing approach, security patterns, CI/CD setup, documentation format. Each wrong choice costs days or weeks to fix later.

This repository gives you those answers immediately - comprehensive, production-tested standards that work together as a complete system. No more arguing about code style. No more wondering about test coverage. No more security reviews finding basic issues.

Just copy, implement, and ship.


πŸ†• Skills System (NEW!)

Last Updated: 2025-10-24 21:45:00 EDT (UTC-04:00)

Load only what you need. Progressive loading reduces token usage by 91-99.6% depending on scenario:

  • Repository metadata: 127K β†’ 500 tokens (99.6% reduction)
  • Typical usage: 8.9K β†’ 573 tokens (93.6% reduction)
  • Single skill Level 1: ~300-600 tokens

Evidence: 61 active skills with progressive disclosure (Level 1: metadata, Level 2: instructions, Level 3: resources)

What Are Skills?

Instead of loading massive 50,000+ token documents, use progressive skills that deliver the right information at the right time:

Note: The @load directive is planned for v2.0. Current implementation uses the skill-loader script.

Current (v1.x):

# Load Level 1: Quick Start (5 minutes, ~336 tokens)
python3 scripts/skill-loader.py load skill:coding-standards

# Load by product type (auto-selects relevant skills)
python3 scripts/skill-loader.py load product:api --language python
# Loads: coding-standards, security-practices, testing, nist-compliance

# Total: ~1,755 tokens compared to loading all standards documents (~150K tokens) (98.8% reduction)

Planned (v2.0):

@load skill:coding-standards
@load product:api --language python

Quick Skills Tutorial (2 minutes)

Current (v1.x):

# 1. Get skill recommendations for your project
python3 scripts/skill-loader.py recommend ./

# 2. Load recommended skills
python3 scripts/skill-loader.py load product:api

# 3. See the difference!
# Before: ~150,000 tokens (full standards)
# After: ~1,755 tokens (Level 1)

Planned (v2.0):

# 2. Load recommended skills
@load product:api

Available Skills (61 Total)

Skill Count Verified: 2025-10-24 19:21:55 EDT (UTC-04:00)

Core Skills (examples with estimated Level 1 token counts):

  • coding-standards: Code quality patterns (~300-400 tokens L1)
  • security-practices: Modern security (~400-500 tokens L1)
  • testing: TDD & testing strategies (~400-500 tokens L1)
  • nist-compliance: NIST 800-53r5 controls (~500-600 tokens L1)

Note: Token counts are estimates. Actual counts vary based on skill complexity. See skills/ directory for complete catalog.

View Full Catalog β†’ | Quick Start Guide β†’ | Skills Directory β†’


⚑ Quick Start for New Projects

Copy these templates to your new repository and let AI do the rest:

Option 1: Full Kickstart Template (Recommended)

# 1. Copy the kickstart template to your new repo
curl -o KICKSTART.md https://raw.githubusercontent.com/williamzujkowski/standards/master/templates/KICKSTART_REPO.md

# 2. Fill in basic project details (or let AI auto-detect)

# 3. Provide KICKSTART.md to any AI assistant (Claude, ChatGPT, Gemini)
# AI will auto-detect your tech stack and generate:
# - Complete project structure
# - Standards-aligned code
# - Configuration files
# - CI/CD pipelines
# - Testing setup
# - Security implementation

Option 2: Use LLM Prompt Directly

  1. Copy this prompt: KICKSTART_PROMPT.md
  2. Paste into any LLM (ChatGPT, Claude, Gemini, etc.) with your project description
  3. Get instant analysis:
    • Complete project structure
    • Relevant standards selection
    • Implementation roadmap
    • Tool configurations
    • First PR ready to go

Example:

You: [Paste kickstart prompt] + "I'm building a Python API with FastAPI and PostgreSQL"

AI: Here's your complete setup:
- Project structure with /src, /tests, /docs
- FastAPI best practices from CODING_STANDARDS.md
- pytest configuration from TESTING_STANDARDS.md
- PostgreSQL patterns from DATA_ENGINEERING_STANDARDS.md
- Docker setup from CLOUD_NATIVE_STANDARDS.md
- GitHub Actions from .github/workflows/
[... complete implementation plan ...]

What You Get

The AI assistant will generate a complete PROJECT_PLAN.md including:

  • Auto-detected tech stack from your repository
  • Standards recommendations using @load directives (v2.0 - planned) or skill-loader script (v1.x - current)
  • Project structure with full directory tree
  • Configuration files (pyproject.toml, package.json, etc.)
  • CI/CD pipelines with quality gates
  • Security implementation with NIST controls
  • Testing strategy with 80%+ coverage targets
  • 8-week implementation timeline with phases

Common Scenarios

Note: The @load directive examples below are planned for v2.0. Current implementation uses skill-loader script.

Python API (v2.0 - planned):

@load [product:api + CS:python + TS:pytest + SEC:* + DE:database]

Python API (v1.x - current):

python3 scripts/skill-loader.py load product:api --language python

React Web App (v2.0 - planned):

@load [product:frontend-web + FE:react + SEC:auth + DOP:ci-cd]

React Web App (v1.x - current):

python3 scripts/skill-loader.py load product:frontend-web --framework react

Mobile App (v2.0 - planned):

@load [product:mobile + CS:swift + TS:xctest + SEC:mobile-auth]

Mobile App (v1.x - current):

python3 scripts/skill-loader.py load product:mobile

Data Pipeline (v2.0 - planned):

@load [product:data-pipeline + CS:python + DE:* + OBS:monitoring]

Data Pipeline (v1.x - current):

python3 scripts/skill-loader.py load product:data-pipeline --language python

πŸ“‹ Available Templates

Core Templates

Template Purpose Usage
KICKSTART_REPO.md LLM-optimized repo kickstart Copy to new repo, fill details, provide to AI
PROJECT_PLAN_TEMPLATE.md Systematic project planning AI generates this from kickstart

Example Projects

Example Tech Stack Located In
Python API FastAPI + PostgreSQL examples/project-templates/python-api/
React SPA React + TypeScript examples/project-templates/react-spa/
NIST-compliant Service Security controls examples/nist-templates/

πŸ”„ For Existing Projects

# Quick assessment
curl -O https://raw.githubusercontent.com/williamzujkowski/standards/master/scripts/setup-project.sh
chmod +x setup-project.sh
./setup-project.sh --assess my-project

# Or manually:
1. Review UNIFIED_STANDARDS.md for gaps
2. Copy relevant templates from examples/
3. Run validation: python scripts/generate-audit-reports.py

πŸ“š What You Get

Complete Standards Library (25 Documents)

Core Development

Specialized Domains

  • Frontend, Mobile, Backend, Data Engineering
  • Cloud Native, DevOps, Observability
  • AI/ML, Event-Driven, Microservices
  • Cost Optimization, Legal Compliance

Ready-to-Use Templates

examples/
β”œβ”€β”€ project-templates/     # Python, JS/TS, Go starter projects
β”œβ”€β”€ nist-templates/        # Security components with compliance tags
β”œβ”€β”€ docker/                # Container configurations
└── ci-cd/                 # GitHub Actions workflows

Automation & Tools

  • Setup Scripts: Auto-configure new projects
  • Validation Tools: Check standards compliance
  • NIST Tagging: Security control annotations
  • VS Code Extension: Real-time compliance hints
  • Pre-commit Hooks: Enforce standards automatically

🎯 For Different Roles

Developers

  • Copy working code patterns
  • Skip bikeshedding discussions
  • Focus on building features
  • Pass code reviews easily

Tech Leads

  • Onboard team members faster
  • Maintain consistency across projects
  • Reduce technical debt
  • Implement best practices systematically

Architects

  • Reference architecture patterns
  • Security-by-default designs
  • Scalability guidelines
  • Integration strategies

Compliance Teams

  • NIST 800-53r5 control templates
  • Audit-ready documentation
  • Automated compliance checking
  • Evidence collection tools

πŸ”₯ Real Examples

Start a Python API

Planned (v2.0):

@load [product:api + CS:python + TS:pytest + SEC:auth]
# Gets you: FastAPI structure, pytest config, JWT auth, Docker, CI/CD

Current (v1.x):

python3 scripts/skill-loader.py load product:api --language python
# Gets you: FastAPI structure, pytest config, JWT auth, Docker, CI/CD

Build a React App

Planned (v2.0):

@load [product:frontend-web + FE:react + SEC:*]
# Gets you: React patterns, testing, all security standards, deployment

Current (v1.x):

python3 scripts/skill-loader.py load product:frontend-web --framework react
# Gets you: React patterns, testing, all security standards, deployment

Data Pipeline

Planned (v2.0):

@load [product:data-pipeline + DE:* + OBS:monitoring]
# Gets you: ETL patterns, data quality, monitoring, orchestration

Current (v1.x):

python3 scripts/skill-loader.py load product:data-pipeline
# Gets you: ETL patterns, data quality, monitoring, orchestration

πŸ“ Repository Structure

standards/
β”œβ”€β”€ docs/
β”‚   β”œβ”€β”€ standards/          # 24 comprehensive standards
β”‚   β”œβ”€β”€ guides/             # Implementation guides
β”‚   └── nist/               # NIST compliance docs
β”œβ”€β”€ examples/               # Copy-paste templates
β”œβ”€β”€ scripts/                # Automation tools
β”œβ”€β”€ config/                 # Configuration files
β”œβ”€β”€ .github/workflows/      # CI/CD templates
└── CLAUDE.md               # LLM interface & routing

βš–οΈ Legal & License

MIT License - Free for commercial use. Standards provided "as-is" without warranty. Legal compliance documents are templates, not legal advice. Always consult professionals for your specific needs.


🀝 Contributing

We actively welcome contributions! The standards evolve with real-world usage.

  • Report issues you encounter
  • Submit PRs with improvements
  • Share templates that work
  • Add examples from your projects

See CREATING_STANDARDS_GUIDE.md for guidelines.

πŸ“– Documentation (MkDocs)

This repository uses MkDocs with the Material theme for documentation.

Local Development:

# Install dependencies
pip install -r requirements.txt

# Serve documentation locally (with live reload)
mkdocs serve
# Visit http://127.0.0.1:8000

# Build static site
mkdocs build

# Build with strict mode (fails on warnings)
mkdocs build --strict

Features:

  • Material theme with light/dark mode
  • Enhanced search functionality
  • Mobile-responsive navigation
  • Automatic deployment via GitHub Actions

Documentation automatically deploys to GitHub Pages when changes are pushed to the master branch.


πŸ€– Claude Code Integration

Make all standards skills and agents globally available in Claude Code via auto-discovery.

Quick Install

# Run from standards repo root
./scripts/sync-to-claude.sh

# Verify installation
ls ~/.claude/skills/std-* | head -5
ls ~/.claude/agents/std-*.md | head -5

What Gets Installed

  • 61 Skills: Symlinked to ~/.claude/skills/std-*
  • 60 Agents: Symlinked to ~/.claude/agents/std-*.md

All items prefixed with std- for easy identification and to avoid conflicts.

Uninstall

./scripts/sync-to-claude.sh --uninstall

🚦 Quick Links

Get Started

Skills (NEW!)

Resources

Support


Stop debating. Start shipping. Your standards are here.

Catalog (auto)

About

LLM Software Development Standards Start any project right in 30 seconds. Battle-tested standards from real production systems.

Resources

License

Stars

Watchers

Forks

Contributors 3

  •  
  •  
  •