Claude Code Configuration for SDLC Projects

A comprehensive .claude configuration for full-stack web development teams. Covers the entire Software Development Life Cycle with specialized rules, skills (slash commands), agents, and automated hooks.

Features

• 12 Specialized Rule Files covering frontend, backend, DevOps, architecture, testing, and more
• 37 Skills (Slash Commands) across 8 categories with agent-team patterns
• 11 Configured Agents for task delegation and specialized workflows
• 10 Production-Ready Hooks for security, quality, and UX automation
• Complete SDLC Coverage for all team roles
• Quality Gates with automated and manual verification checklists

Tech Stack

Category	Technologies
Frontend	React 18+, Material-UI v5+
Backend	Node.js, Express.js
Database	PostgreSQL
Testing	Jest/Vitest, Playwright, React Testing Library
CI/CD	GitHub Actions
Containers	Docker, Docker Compose
Validation	Zod

Structure

.claude/
├── CLAUDE.md                          # Main configuration
├── settings.json                      # Hooks, permissions, agents
├── rules/
│   ├── api-design.md                  # REST conventions, OpenAPI
│   ├── architecture.md                # ADRs, system design, Mermaid
│   ├── backend.md                     # Node.js/Express patterns
│   ├── code-review.md                 # PR templates, review checklists
│   ├── database.md                    # PostgreSQL schema, migrations
│   ├── devops.md                      # GitHub Actions, Docker
│   ├── documentation.md               # Technical writing, runbooks
│   ├── frontend.md                    # React/MUI patterns
│   ├── no-ascii-diagrams.md           # Enforce Mermaid over ASCII
│   ├── quality-gates.md               # QA processes, release criteria
│   ├── security.md                    # OWASP, auth, validation
│   └── testing.md                     # Unit, integration, E2E
└── skills/
    ├── architecture/                  # 8 skills
    ├── diagramming/                   # 3 skills
    ├── content-processing/            # 8 skills
    ├── codebase-health/               # 6 skills
    ├── scoring/                       # 2 skills
    ├── reporting/                     # 2 skills
    ├── meetings/                      # 3 skills
    └── knowledge/                     # 5 skills
hooks/
├── security/
│   ├── secret-detection.py            # Block secrets in prompts
│   ├── secret-file-scanner.py         # Block secrets in file writes
│   └── file-protection.py            # Protect .env, lockfiles, keys
├── quality/
│   ├── code-formatter.py             # Auto-format after edits
│   ├── import-checker.py             # Validate import paths
│   └── naming-convention.py          # Enforce file naming
├── ux/
│   ├── context-loader.sh             # Auto-load skill context
│   └── search-hint.sh               # Suggest faster searches
├── safety/
│   └── bash-safety.py               # Auto-allow safe commands
└── notification/
    └── desktop-notify.sh             # Desktop notification on completion
docs/
└── agent-teams-guide.md              # Agent team patterns guide

Quick Start

Option 1: Clone and Copy

git clone https://github.com/servika/claude-configuration.git
cp -r claude-configuration/.claude your-project/
cp -r claude-configuration/hooks your-project/

Option 2: Use degit

npx degit servika/claude-configuration your-project-config
cp -r your-project-config/.claude your-project/
cp -r your-project-config/hooks your-project/

Hook Setup

After copying, make hook scripts executable:

chmod +x hooks/**/*.py hooks/**/*.sh

Hooks are configured in .claude/settings.json and run automatically during Claude Code sessions.

SDLC Role Coverage

Role	Primary Rules	Focus Areas
Developer	frontend, backend, api-design, database	Feature implementation, code quality
QA Engineer	testing, quality-gates	Test automation, E2E, release criteria
Technical Writer	documentation, architecture	Docs, runbooks, ADRs
DevOps Engineer	devops, security	CI/CD, Docker, deployment
Architect	architecture, api-design, database	System design, technical decisions

Configured Agents

Agent	Description	Key Rules
frontend	React + MUI development	frontend, testing, security
backend	Node.js + Express APIs	backend, api-design, database, security
devops	CI/CD and deployment	devops, security
architect	System design and ADRs	architecture, api-design, database
qa	E2E testing and quality	testing, quality-gates
api-designer	REST API design	api-design, documentation
documentation	Technical writing	documentation, architecture
testing	All test types	testing, quality-gates
code-review	PR quality assessment	code-review, security
security	Security review	security, backend
database	Schema and queries	database, backend

Skills (Slash Commands)

37 skills organized across 8 categories. Skills use agent-team patterns (fan-out/fan-in, batch, triage) for parallel analysis.

Architecture

Command	Description
/adr	Generate Architecture Decision Records
/impact-analysis	4-agent impact analysis (technical, organisational, financial, risk)
/scenario-compare	Compare architectural scenarios with weighted scoring
/nfr-capture	Capture Non-Functional Requirements (ISO 25010)
/nfr-review	Review NFRs for completeness, measurability, feasibility
/architecture-report	5-agent governance report with RAG health dashboard
/cost-analysis	Infrastructure, licensing, and operational cost analysis
/dependency-graph	Generate Mermaid dependency graphs with criticality

Diagramming

Command	Description
/diagram	Generate Mermaid diagrams (C4, data flow, sequence, ER)
/c4-diagram	C4-specific diagrams (Context, Container, Component)
/diagram-review	4-agent review for readability, completeness, accuracy

Content Processing

Command	Description
/pdf-extract	Extract and structure content from PDFs
/pptx-extract	Extract content from PowerPoint files
/youtube-analyze	Analyse YouTube videos (technical/summary/action-items)
/video-digest	Triage and deep-analyse video content
/weblink	Quick web page capture with summary
/article	Article capture with relevance scoring
/research-notes	Structured notes from research materials
/document-extract	Auto-detect and extract from any document format

Codebase Health

Command	Description
/code-quality-report	5-dimension quality analysis (complexity, coverage, lint, deps, build)
/broken-references	Find broken imports, missing files, dead exports
/dead-code-finder	Detect unused files, unreachable components, dead routes
/auto-document	Batch-generate JSDoc, README sections, API docs
/auto-categorize	Classify files by architectural layer
/dependency-checker	Audit npm deps for security, freshness, usage, licenses

Scoring

Command	Description
/score-document	Score documents against rubrics with parallel agents
/exec-summary	Generate audience-tailored executive summaries (BLUF)

Reporting

Command	Description
/sprint-summary	Generate sprint summaries from git history and PRs
/project-report	Project status report with RAG indicators

Meetings

Command	Description
/meeting-notes	Structure meeting notes with parallel extraction
/voice-meeting	Process voice transcripts into structured notes
/email-capture	Capture emails as structured notes with action items

Knowledge

Command	Description
/summarize	Summarise content at configurable depth and audience
/find-related	Discover related code via imports, deps, patterns
/find-decisions	Catalogue decisions from ADRs, PRs, commits
/timeline	Generate timelines from git history and milestones
/skill-creator	Create new skills with agent-team patterns

Hooks

Hooks run automatically during Claude Code sessions to enforce quality and security.

Hook	Type	Purpose
secret-detection.py	UserPromptSubmit	Block secrets (AWS, JWT, Stripe, etc.) in prompts
secret-file-scanner.py	PreToolUse	Block secrets in file writes
file-protection.py	PreToolUse	Protect .env, lockfiles, keys from modification
code-formatter.py	PostToolUse	Auto-format with Prettier/Black/gofmt after edits
import-checker.py	PostToolUse	Validate JS/TS import paths resolve to files
naming-convention.py	PostToolUse	Enforce file naming conventions per directory
context-loader.sh	UserPromptSubmit	Auto-load relevant rules when skills are invoked
search-hint.sh	PreToolUse	Suggest faster search methods (Glob over Grep)
bash-safety.py	PreToolUse	Auto-allow safe read-only commands
desktop-notify.sh	Stop	Desktop notification (macOS/Linux) on completion

Hook Lifecycle

UserPromptSubmit → PreToolUse → [Tool Executes] → PostToolUse → Stop

Exit codes: 0 = allow, 1 = warn (message shown), 2 = block (tool prevented)

What's Included

Development Rules

• frontend.md - React component patterns, MUI styling with sx prop, hooks, forms, accessibility
• backend.md - Express architecture (routes → controllers → services), error handling, validation, auth
• api-design.md - REST conventions, HTTP methods/status codes, pagination, versioning, OpenAPI specs
• database.md - PostgreSQL schema design, naming conventions, migrations, query optimization

Quality Rules

• testing.md - Unit testing with Jest, integration with Supertest, E2E with Playwright, Page Object Model
• quality-gates.md - PR checks, CI pipeline gates, staging QA, release criteria, performance thresholds
• code-review.md - PR templates, review checklists, comment etiquette, approval criteria
• security.md - OWASP Top 10 prevention, input validation, JWT security, rate limiting, XSS prevention

Operations Rules

• devops.md - GitHub Actions workflows, Docker configuration, deployment strategies, health checks
• architecture.md - ADR templates, system design documentation, Mermaid diagrams, technical debt tracking
• documentation.md - README structure, API docs, runbooks, incident response templates
• no-ascii-diagrams.md - Enforce Mermaid diagrams, never ASCII art

Code Quality Standards

Automated Verification

• ESLint must pass
• Build must succeed
• Test coverage ≥60% overall, ≥20% per file
• Security audit (npm audit) clean

Code Requirements

• JSDoc comments on all functions
• Parameterized database queries (no SQL injection)
• Input validation with Zod
• No console.log in production code

Customization

Modify Tech Stack

Edit .claude/CLAUDE.md to change frameworks, libraries, or coverage thresholds.

Add Custom Rules

Create new files in .claude/rules/ and reference them in settings.json under the appropriate agent.

Add Custom Skills

Create new files in .claude/skills/<category>/ with YAML frontmatter:

---
description: "Short description for /my-skill"
model: sonnet
---

# My Skill

Instructions for the skill...

Or use /skill-creator to generate skills interactively.

Adjust Permissions

Edit .claude/settings.json to add or remove allowed/denied commands.

Customise Hooks

Hook scripts in hooks/ can be modified to adjust patterns, thresholds, or behaviour. Key customisation points:

• secret-detection.py — Add patterns to SECRET_PATTERNS list
• file-protection.py — Modify PROTECTED_PATHS and SAFE_DIRECTORIES
• bash-safety.py — Extend SAFE_COMMANDS or NEVER_SAFE lists
• naming-convention.py — Adjust CONVENTIONS per directory

Requirements

• Claude Code or compatible tool
• Node.js 18+ (for the configured tech stack)
• Python 3 (for hook scripts)
• PostgreSQL 14+ (if using database rules)
• Docker (if using DevOps rules)

Contributing

Contributions are welcome! Please:

1. Fork the repository
2. Create a feature branch
3. Make your changes
4. Submit a pull request

Guidelines

• Follow the existing rule file format
• Include practical code examples
• Add checklists where appropriate
• Test with Claude Code before submitting

Acknowledgments

• Built for teams using Claude Code who want consistent, high-quality development practices across the entire SDLC
• Skills adapted from Dave's Claude Code Skills with significant adaptation for SDLC/web development workflows

License

MIT License - see LICENSE for details.