Skip to content

nobelk/claude-tools

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

24 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Claude Code Skills & Prompts

License

A curated collection of Claude Code skills and prompt templates designed to streamline software engineering workflows. These skills leverage Claude's advanced capabilities to provide automated code analysis, design reviews, security scanning, and comprehensive documentation generation.

Installation

  1. Clone or download this repository:

    git clone https://github.com/yourusername/claude-tools.git
  2. Copy skills to your Claude Code settings (Option A - Per Project):

    # Copy the skills directory to your project
    cp -r claude-tools/skills /path/to/your/project/.claude/
  3. Or use as a global skills library (Option B):

    # Copy to your global Claude Code directory
    cp -r claude-tools/skills ~/.claude/
  4. Verify installation:

    • Open your project in Claude Code
    • The skills should be available based on context triggers or explicit invocation

Available Skills

Code Review & Quality

code-review - Pull Request Analysis

Directory: skills/code-review/

Performs senior-engineer-grade PR code reviews covering correctness, security, performance, testing, design, and observability.

Triggers: Review a pull request, PR, merge request, diff, branch, or code changes

Key Features:

  • Multi-dimensional analysis (correctness, security, performance, design)
  • File prioritization by risk (P0-P4)
  • Severity classification (CRITICAL/HIGH/MEDIUM/LOW/PRAISE)
  • Actionable recommendations with code suggestions

codebase-review - Codebase Audit

Directory: skills/codebase-review/

Performs a thorough codebase review and produces a prioritized to_do.md listing all bug fixes, improvements, and action items.

Triggers: Review, audit, or analyze a codebase for bugs, code quality issues, missing tests, TODOs, technical debt

Key Features:

  • Systematic file-by-file review
  • Bug detection (logical, security, performance)
  • Test coverage gap identification
  • .NET/C#-specific checks included
  • Prioritized action item list

dotnet-refactor - .NET Refactoring

Directory: skills/dotnet-refactor/

Principal-level C#/.NET codebase refactoring with phased plans and regression safety.

Triggers: Refactor, restructure, clean up, or improve C#/.NET code quality

Key Features:

  • Discovery of code smells and coupling hotspots
  • Baseline unit test safety net
  • Phased refactoring plan generation
  • SOLID principles and design pattern guidance
  • Logging and cross-cutting concern improvements

Security & Performance

security-scan - Vulnerability Detection

Directory: skills/security-scan/

Comprehensive security vulnerability assessment aligned with OWASP, CWE, and CVSS standards.

Triggers: Scan code for security issues, perform security review/audit, find vulnerabilities, check for OWASP Top 10 issues

Key Features:

  • OWASP Top 10 vulnerability detection
  • Multi-language support (JS/TS, Python, Java, Go, C/C++, Ruby, PHP, C#, Rust, Swift, Kotlin)
  • Automated pattern scanning with manual deep-dive
  • Dependency audit for known CVEs
  • CVSS-aligned severity classification

performance-analysis - Performance Profiling

Directory: skills/performance-analysis/

Analyzes application performance bottlenecks with focus on .NET diagnostics.

Triggers: Performance analysis, profiling, optimization recommendations

Key Features:

  • Anti-pattern detection
  • .NET-specific diagnostics
  • Structured output template
  • Actionable optimization recommendations

Architecture & Design

architecture-review - Architecture Analysis

Directory: skills/architecture-review/

Generates industry-standard architecture design documents by analyzing codebases across structure, patterns, principles, and quality attributes.

Triggers: Review or analyze code architecture, create architecture overview, assess code quality, identify architectural strengths/weaknesses

Key Features:

  • Systematic codebase exploration
  • Design pattern identification (creational, structural, behavioral, architectural)
  • Clean code principles assessment (SOLID, DRY, KISS, YAGNI)
  • Quality attribute ratings (1-5 scale)
  • Mermaid diagram generation

design-doc-review - Design Document Review

Directory: skills/design-doc-review/

Principal-engineer-level design document review against PRD requirements.

Triggers: Review, critique, or evaluate a design document, technical design, architecture proposal

Key Features:

  • PRD alignment verification
  • 11 review dimensions (architecture, security, scalability, availability, etc.)
  • Severity-based issue classification (P0-P3)
  • Actionable recommendations

prd-to-design-doc - PRD to Design Document

Directory: skills/prd-to-design-doc/

Transforms Product Requirements Documents into Developer Design Documents following RFC conventions used by Google, Uber, Airbnb, Stripe, and others.

Triggers: Create a design document from a PRD, generate an RFC, create technical design doc

Key Features:

  • Trade-off focused analysis
  • Alternatives considered with comparison
  • Cross-cutting concerns (security, privacy, observability)
  • Phased rollout planning
  • Mermaid architecture diagrams

Documentation & Requirements

prd-generator / code-to-prd-doc - PRD from Codebase

Directory: skills/prd-generator/ | skills/code-to-prd-doc/

Reverse-engineers a codebase into a formal Product Requirements Document structured for Jira work item creation.

Triggers: Generate PRD, analyze codebase, document requirements, reverse-engineer features, prepare Jira epics from code

Key Features:

  • Feature extraction by category
  • Gap and technical debt identification
  • Jira-ready epic/story/task breakdown
  • .NET/C# specific analysis patterns
  • Fibonacci story point estimation

Testing & Bug Fixing

tdd-bugfix - Test-Driven Bug Fixing

Directory: skills/tdd-bugfix/

Fixes bugs using test-driven development methodology with strict adherence to project conventions.

Triggers: Fix a bug using TDD, safe bug fix with tests

Key Features:

  • Project convention discovery
  • Failing test creation before fix
  • Project test framework/style matching
  • Full regression test verification

qa-test-plan-review - Railway QA Test Plan Review

Directory: skills/qa-test-plan-review/

Railway QA Lead Engineer skill for reviewing test plans against CENELEC EN 50716/EN 50128, EN 50126, and EN 50129 standards.

Triggers: Railway test plan review, CENELEC compliance testing, signaling test coverage, safety-critical test case generation

Key Features:

  • SIL-graded requirement analysis
  • Gap analysis against railway standards
  • CSV test case generation matching input format
  • Safety/fail-safe scenario coverage

Utilities

git-rebase-main - Git Rebase Helper

Directory: skills/git-rebase-main/

Fetches and rebases current branch onto main/master with automatic stash handling.

Triggers: Update branch with main, sync with default branch, pull and rebase

Key Features:

  • Auto-detects main vs master
  • Automatic stash/unstash via --autostash
  • Conflict resolution guidance

keybindings-help - Keyboard Shortcuts Help

Directory: skills/keybindings-help/

Helps customize Claude Code keyboard shortcuts by editing ~/.claude/keybindings.json.

Triggers: Rebind keys, add chord bindings, customize keybindings, change submit key

Key Features:

  • Complete keybindings.json reference
  • All contexts and actions documented
  • Chord binding support
  • Common customization examples

Prompt Templates

Reusable prompt templates in the prompts/ directory:

File Purpose
analyze_paper.md Academic paper analysis template
architecture_analysis.md Comprehensive codebase architecture review
design_document.md Design document generation template
engineering_brief.md Daily engineering brief from Jira data

Repository Structure

claude-tools/
├── skills/                          # Claude Code skills
│   ├── architecture-review/
│   │   ├── SKILL.md
│   │   └── references/
│   ├── code-review/
│   │   ├── SKILL.md
│   │   └── references/
│   ├── code-to-prd-doc/
│   │   ├── SKILL.md
│   │   ├── assets/
│   │   └── references/
│   ├── codebase-review/
│   │   └── SKILL.md
│   ├── design-doc-review/
│   │   ├── SKILL.md
│   │   └── references/
│   ├── dotnet-refactor/
│   │   ├── SKILL.md
│   │   └── references/
│   ├── git-rebase-main/
│   │   ├── SKILL.md
│   │   └── scripts/
│   ├── keybindings-help/
│   │   └── SKILL.md
│   ├── performance-analysis/
│   │   ├── SKILL.md
│   │   ├── references/
│   │   └── scripts/
│   ├── prd-generator/
│   │   ├── SKILL.md
│   │   ├── assets/
│   │   └── references/
│   ├── prd-to-design-doc/
│   │   └── SKILL.md
│   ├── qa-test-plan-review/
│   │   ├── SKILL.md
│   │   └── references/
│   ├── security-scan/
│   │   ├── SKILL.md
│   │   ├── references/
│   │   └── scripts/
│   └── tdd-bugfix/
│       └── SKILL.md
├── prompts/                         # Reusable prompt templates
│   ├── analyze_paper.md
│   ├── architecture_analysis.md
│   ├── design_document.md
│   └── engineering_brief.md
├── workflow/                        # Workflow documentation & assets
├── LICENSE
└── README.md

Skill Structure

Each skill follows a consistent structure:

skill-name/
├── SKILL.md          # Main skill definition with frontmatter and workflow
├── references/       # Supporting documentation and checklists
├── assets/           # Templates and output formats
└── scripts/          # Automation scripts (if applicable)

The SKILL.md file contains:

  • Frontmatter: Name, description, and trigger conditions
  • Workflow: Step-by-step execution process
  • References: Links to supporting materials

Benefits

  • Productivity: Automate time-consuming analysis and documentation tasks
  • Consistency: Standardized analysis approaches across all scenarios
  • Thoroughness: Comprehensive coverage of quality, security, and performance dimensions
  • Actionability: Prioritized recommendations with specific code examples
  • Collaboration: Structured outputs perfect for team reviews and discussions
  • Best Practices: Incorporates industry standards (OWASP, SOLID, CENELEC, etc.)

Usage Tips

  1. Context Triggers: Skills activate automatically based on your request context
  2. Reference Files: Skills use supporting references for comprehensive analysis
  3. Customization: Edit SKILL.md files to adapt to your team's specific needs
  4. Iteration: Skills are designed for iterative refinement based on initial output
  5. Headless/CI Automation: Run the full agentic workflow non-interactively.
# Run the full workflow non-interactively
claude -p "
  Implement rate limiting middleware on the ASP.NET Core API using
  System.Threading.RateLimiting. Apply a fixed-window policy of
  100 requests per minute per client IP on all /api/* endpoints.
  Follow this workflow:
  1. Use software-architect agent to produce implementation plan.
  2. Use principal-engineer agent to review plan and implement with tests.
  3. Use code-reviewer agent to review and fix all issues. If BLOCKED,
     use principal-engineer to rework, then re-run code-reviewer.
  4. Use test-runner agent to run all four test tiers.
  5. If any tier fails, use principal-engineer to fix, then code-reviewer
     to verify, then test-runner to re-test. Repeat until green.
  Print WORKFLOW_SUCCESS if everything passes, WORKFLOW_FAILED otherwise.
" --allowedTools "Read,Write,Edit,Bash,Grep,Glob,MultiEdit" --output-format json

Multiagent Workflow Prompt

I need to add rate limiting middleware to the ASP.NET Core API using the built-in System.Threading.RateLimiting. Apply a fixed-window policy of 100 requests per minute per client IP on all /api/* endpoints. Please follow this workflow:

  1. Use the software-architect agent to analyze the codebase and produce a detailed implementation plan with unit test strategy.

  2. Use the principal-engineer agent to review and refine the plan, then implement the code changes with unit tests. Unit tests must pass.

  3. Use the code-reviewer agent to review the implementation and fix any issues directly. Build and unit tests must pass after fixes. If the verdict is BLOCKED, use the principal-engineer agent to rework the implementation, then re-run the code-reviewer agent.

  4. Use the test-runner agent to run all four test tiers (unit, functional, integration, pact).

  5. If any tests fail due to the change, use the principal-engineer agent to fix them, then the code-reviewer agent to verify the fixes, then re-run the test-runner agent. Repeat until all tiers pass.

  6. Summarize the final result.

  7. Use principal-engineer agent to review plan and implement with tests.

  8. Use code-reviewer agent to review and fix all issues. If BLOCKED, use principal-engineer to rework, then re-run code-reviewer.

  9. Use test-runner agent to run all four test tiers.

  10. If any tier fails, use principal-engineer to fix, then code-reviewer to verify, then test-runner to re-test. Repeat until green. Print WORKFLOW_SUCCESS if everything passes, WORKFLOW_FAILED otherwise.

Contributing

We welcome contributions! When adding new skills:

  1. Follow the structure: Use the established SKILL.md format with frontmatter
  2. Be comprehensive: Include detailed workflow steps and references
  3. Provide clear triggers: Document what user requests activate the skill
  4. Include references: Add supporting documentation in a references/ subdirectory
  5. Test thoroughly: Validate skills with real-world scenarios
  6. Update README: Add documentation for new skills to this file

License

This project is licensed under the Apache License 2.0 - see the LICENSE file for details.

Support


Made for the engineering community

About

Tested Claude commands for software engineering tasks

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors