Reflexive-Claude-Code

🥈Silver

A skills-driven agentic context engineering workflow for Claude Code that structures AI agent behavior through composable skills, rules, and CLAUDE.md guidance.

1620Updated 1mo ago

Intermediate30min to implementautomation

Saves ~80 min per use

Quick InstallView Source

git clone https://github.com/wayne930242/Reflexive-Claude-Code.git

Works with:

Claude

Overview

About This Skill

Reflexive-Claude-Code is a structured workflow system for Claude Code that organizes AI agent behavior through composable skills, project rules, and CLAUDE.md configuration. It provides a plugin-based architecture under .claude-plugin that includes skill authoring, subagent planning, parallel agent dispatch, and reflexive reviewing capabilities. The system addresses common context engineering challenges such as CLAUDE.md bloat, hard-rule compliance gaps, and agent orchestration by offering templates, reviewers, and guidance grounded in 2026 prompting research. It supports monorepo setups with nested CLAUDE.md files, conditional instruction dispatch, and a safety bypass prevention system for destructive operations. Developers building complex agentic workflows with Claude Code benefit from the structured scaffolding and opinionated best practices this skill provides.

How to Use

Install the plugin by adding the .claude-plugin directory to your Claude Code project. Configure your orchestrator model and settings in .rcc/config.yml, which records migration state, model preferences, and safety scope decisions. Use the provided skills for writing rules, authoring subagents, and reviewing CLAUDE.md files. Release automation via release-please can be set up through the creating-plugins workflow.

Use Cases

Structuring Claude Code agent behavior with composable, reviewable skills

Dispatching parallel subagents for large or multi-step development tasks

Enforcing safety rules against destructive operations like force pushes or test deletion

Authoring and validating CLAUDE.md files to avoid context bloat and improve agent performance

Setup & Installation

Quick Install

No install command available. Check the GitHub repository for manual installation instructions.

Alternative Install (Git Clone)

git clone https://github.com/wayne930242/Reflexive-Claude-Code

Requirements

Claude Code or compatible AI agent
Works with: Claude

Quick Start Guide

Install the Skill

Copy the install command above and run it in your terminal.

Open Your AI Agent

Launch Claude Code, Cursor, or your preferred AI coding agent.

Try It Out

Use the prompt template or examples below to test the skill.

Customize

Adapt the skill to your specific use case and workflow.

Usage Examples

Prompt Template

I'm building a [PROJECT_TYPE] for [COMPANY] in the [INDUSTRY] sector. I need to create a workflow that automates [SPECIFIC_TASK]. Here's the context: [CONTEXT]. Can you generate a skills-driven Agentic Context Engineering workflow for Claude Code that includes: 1) Skill identification, 2) Context engineering, 3) Agentic execution, and 4) Continuous improvement. Make it specific to this use case.

Example Output

## Skills-Driven Agentic Context Engineering Workflow for [PROJECT_TYPE] Automation

### 1. Skill Identification
- **Primary Skills**: Data processing, API integration, error handling
- **Supporting Skills**: Data validation, logging, notification systems
- **Contextual Skills**: Industry-specific compliance, data security protocols

### 2. Context Engineering
- **Initial Context**: [CONTEXT]
- **Expanded Context**:
  - Data sources: [DATA_SOURCE_1], [DATA_SOURCE_2]
  - Integration points: [API_1], [API_2]
  - Compliance requirements: [REGULATION_1], [REGULATION_2]

### 3. Agentic Execution
- **Agent 1**: Data Collection Agent
  - Responsibilities: Gather data from [DATA_SOURCE_1] and [DATA_SOURCE_2]
  - Tools: Python scripts, API connectors
- **Agent 2**: Data Processing Agent
  - Responsibilities: Clean, validate, and transform data
  - Tools: Pandas, custom validation scripts
- **Agent 3**: Integration Agent
  - Responsibilities: Integrate processed data with [API_1] and [API_2]
  - Tools: API clients, middleware

### 4. Continuous Improvement
- **Feedback Loop**: Implement a feedback mechanism to capture user inputs and system errors
- **Iterative Updates**: Schedule regular updates based on feedback and system performance metrics
- **Documentation**: Maintain up-to-date documentation for all agents and workflows