testing-process

This is a starter template repository designed to provide a complete development environment for Claude-Code with pre-configured MCP servers and tools for AI-powered, collaborative, development workflows. The defaults are intentionally minimal, containing only configuration templates for three primary systems: Claude Code, Serena, and Task Master. Users can opt-in to additional claude-code features like [skills](https://code.claude.com/docs/en/skills), [plugins](https://code.claude.com/docs/en/plugins), [hooks](https://code.claude.com/docs/en/hooks-guide), [sub-agents](https://code.claude.com/docs/en/sub-agents), and so on.

> [!NOTE]

> This configuration also focuses on collaborative development workflows where multiple developers are working on the same code-base, which is one of the reasons why most of claude- and mcp-related settings are local-scoped (i.e. most claude settings will be in .claude/settings.local.json so they can be shared with the entire dev team, and not in user-scoped ~/claude/settings.json, which are harder to share with others.)

>

> For this same reason, most of the claude/mcp configuration files are not git-ignored, but instead committed to the repo.

• 🤖 Four Pre-Configured MCP Servers

- Context7: Up-to-date library documentation and code examples

- Serena: Semantic code analysis with LSP integration for intelligent navigation

- Task Master: AI-powered task management and workflow orchestration

- Pal: Multi-model AI integration for debugging, code review, and planning

• ⚙️ Automated Template Cleanup

- GitHub Actions workflow for one-click repository initialization

- Configurable inputs for language detection and Task Master settings

- Automatic cleanup of template-specific files for a clean starting point

• 📋 50+ Task Master Slash Commands

- Pre-configured hierarchical command structure under /project:tm/

- Commands for task management, complexity analysis, PRD parsing, and workflows

- Complete command reference in .claude/TM_COMMANDS_GUIDE.md

• 🔍 Intelligent Code Navigation

- Serena's symbol-based code analysis for efficient exploration

- Token-efficient reading with overview and targeted symbol queries

- Reference tracking and semantic understanding across your codebase

• 📝 Configuration Templates

- Ready-to-use templates for .serena/, .taskmaster/, and .claude/ directories

- Placeholder-based customization with repository-specific values

- Permission configuration for tool access control

• 📚 Comprehensive Documentation

- Project-level CLAUDE.md with integration guidance

- Task Master integration guide with 400+ lines of best practices

- Complete workflow specification and command references

You will need the following on your workstation:

Tools

• [npm](https://www.npmjs.com/package/npm)
• [uv](https://docs.astral.sh/uv/)

API Keys

• [Context7](https://context7.com/) API key
• Gemini API key for [pal-mcp-server](https://github.com/BeehiveInnovations/pal-mcp-server). You don't need to use gemini and can configure pal with any other provider/models. See [pal getting started docs](https://github.com/BeehiveInnovations/pal-mcp-server/blob/main/docs/getting-started.md) for more details.

Claude `claude.json` mcp settings

You need to have mcpServers present and configured in your ~/.claude.json.

> [!NOTE]

> The reason we put them in the user's claude.json configuration, instead of repo local settings, is to prevent committing API keys, which some MCP servers might require.

>

> These configs are also generic enough that they can be re-used across every project, and hence is better placed in user's settings.

Here's an example `mcpSer