zebbern/claude-code-guide
Claude Code Guide
This project provides a framework for AI agent orchestration and context management, enabling the deployment of specialized AI personas and subagents to solve multi-step technical goals. It centers on managing specialized agents with isolated contexts and role-based prompts to handle domain-specific tasks.
The system differentiates itself through a hierarchical project memory using markdown files to maintain coding standards and a secure execution model that utilizes sandboxed environments and git worktree isolation. It also features a Model Context Protocol integration for external tool connectivity and a dedicated browser extension for web automation and UI verification.
The capability surface covers automated code refactoring, cloud-based code review, and developer workflow automation through scheduled jobs and event-driven lifecycle hooks. It includes tools for monitoring token consumption, optimizing prompt caching, and managing session states through automated context persistence.
Features
- AI Agent Orchestration - Coordinates specialized AI personas and subagents to solve complex, multi-step technical goals.
- Agent Workflow Orchestrations - Sequences and coordinates specialized AI agents to solve large technical tasks through high-level orchestration.
- Custom-Prompted Agents - Allows the definition of focused AI assistants with custom system prompts and tool restrictions for targeted tasks.
- Persistent Context Management - Organizes project memory and session states using structured files for consistent coding standards.
- AI Execution Sandboxes - Implements isolated environments specifically designed for running AI coding agents securely.
- MCP Protocol Integrations - Connects to external protocol servers through the Model Context Protocol to provide agents with specialized tools.
- Markdown-Based Project Memory - Uses a hierarchical system of markdown files to maintain shared instructions and coding standards.
- Memory Hierarchies - Organizes project standards and shared instructions in a hierarchical structure of markdown files for consistent context retrieval.
- System Prompt Configurations - Applies pre-defined system prompt configurations to switch agent behavior between roles like auditing or architecture.
- Subagent Orchestrations - Provides the ability to deploy specialized helper agents with isolated contexts to decompose and solve complex technical tasks.
- Development Workflow Automation - Provides systems to manage end-to-end coding processes by integrating terminal access and automated refactoring tools.
- Worktree Isolation Tools - Starts sessions in temporary git worktrees to prevent uncommitted changes from affecting the main branch.
- Interactive REPLs - Provides a terminal-based REPL for interactively analyzing and modifying codebases with AI.
- Sandboxed Execution Environments - Runs shell commands and tool operations in isolated environments or temporary git worktrees to protect the host filesystem.
- Sandboxed Shell Executions - Runs shell commands in isolated environments with sandboxing to prevent accidental filesystem modifications.
- Agentic Workflow Automations - Coordinates multiple background agents to execute multi-step operational processes like system migrations.
- Tool Access Controls - Controls available tool access through allow-lists and deny-lists to ensure safe agent operation.
- AI Personas - Provides a library of reusable behavioral personas and system prompts for roles like architects and auditors.
- LLM Session State Management - Resumes, forks, or clears AI conversations to optimize context window usage.
- Tool Execution Permissions - Defines granular access rules for specific tools and paths to control exactly what the agent can execute.
- AI Agent Role Assignments - Assigns functional roles like architect or developer to AI agents to improve accuracy in specialized domains.
- Documentation-Based Context Persistence - Saves and retrieves context automatically in memory files to maintain knowledge across sessions.
- Hierarchical Subagent Orchestrations - Implements an architectural framework for managing multi-level agent delegation and task decomposition.
- Agent Goal Definitions - Sets specific completion conditions that an agent pursues across multiple interactions until the objective is met.
- Technical Task Skillsets - Uses predefined configurations and prompts for complex tasks like security auditing and refactoring.
- MCP Server Integrations - Integrates with Model Context Protocol servers to provide agents with specialized tools and resources.
- External Tool Integrations - Connects AI assistants to external services, databases, and APIs via standardized protocols.
- Prompt Caching - Optimizes performance and reduces costs by caching system prompts and conversation history.
- Markdown Configurations - Defines AI agent behaviors using markdown files in dedicated directories to automate technical tasks.
- Code Quality and Analysis - Analyzes codebases for correctness and cleanup to enforce coding standards.
- Automated Code Refactoring - Analyzes codebases and applies direct structural modifications through an AI-powered interface.
- Workflow Convention Documentation - Documents workflow rules and quality practices to ensure consistent output across the project.
- Background Command Execution - Executes long-running shell commands asynchronously without blocking the primary interactive interface.
- Process Namespace Isolation - Runs commands in a sandbox with namespace isolation and network restrictions to prevent unauthorized system access.
- Asynchronous Background Processors - Processes long-running commands and AI sessions in non-blocking threads to keep the interactive interface responsive.
- Token Usage Analysis - Breaks down token consumption and identifies context-heavy tools to extend session longevity.
- Agentic Code Reviews - Implements multi-agent analysis and critique for auditing branches and pull requests.
- AI-Driven - Controls web browsers via extensions to perform automated testing and UI verification using AI.
- Agent Browser Controls - Directs a connected web browser to perform actions and extract information through a dedicated browser extension.
- Browser Automation - Provides programmatic control of browser instances for scraping and UI verification.
