🧠 bless_claude_token

An advanced, token-economizing System Prompt for Claude to radically optimize API costs, context limits, and multi-AI task delegation.

English | Español | Français | Deutsch | Português | 中文

---

📖 Overview

bless_claude_token is an engineered System Prompt that transforms Claude into a Senior Software Architect and strict Token Economist. It is designed to aggressively protect your context window, preventing "hallucinations" and saving up to 80% on output tokens.

✨ Core Features

• 🛡️ Strict Diff Patching: Outputs only exact Git-style diff patches for modified lines.
• 🤖 Multi-AI Auto-Delegation: Writes exact prompts for you to copy/paste into faster models (Llama 3/Gemini) for boilerplate.
• 💾 State Hibernation (/save): Generates a Markdown mental map of your project's architecture.
• 🔇 Anti-Noise Filter: Automatically stops processing massive error logs.

⚙️ Installation

1. Navigate to prompt_maestro.md.
2. Copy the raw text.
3. Paste it into the "System Prompt" or "Project Instructions" of your LLM client.

For backend integration, see api_injection_example.js.

📄 License

This project is MIT licensed.