claude-code-system-prompts

🥇Gold

All parts of Claude Code's system prompt, 18 builtin tool descriptions, sub agent prompts (Plan/Explore/Task), utility prompts (CLAUDE.md, compact, statusline, magic docs, WebFetch, Bash cmd, security review, agent creation). Updated for each Claude Code version.

9,0005610Updated 1w ago

Intermediate30min to implementautomation

Saves ~210 min per use

Quick InstallView Source

git clone https://github.com/Piebald-AI/claude-code-system-prompts.git

Works with:

Claude

Overview

About This Skill

https://piebald.ai

How to Use

[{"step":"Identify the utility prompt type needed (e.g., CLAUDE.md, WebFetch, Bash cmd).","action":"Review the 18 builtin tool descriptions in Claude Code's system prompts to match your goal with the right utility.","tip":"Use the `compact` utility prompt for quick, single-command workflows, or `statusline` for real-time progress tracking."},{"step":"Customize the prompt template for your project.","action":"Replace [UTILITY_PROMPT_NAME], [PROJECT_NAME], [SPECIFIC_GOAL], and [RELEVANT_SECTIONS] with your project's details. Include project-specific tools or commands.","tip":"For sub-agent prompts (Plan/Explore/Task), define clear success criteria and error-handling steps."},{"step":"Test the utility prompt in a sandbox environment.","action":"Run the prompt with `claude \"[YOUR_CUSTOM_PROMPT]\"` and verify outputs against expected results. Use `--debug` for troubleshooting.","tip":"For security reviews, include a `security_checklist.md` section in your utility prompt to validate inputs/outputs."},{"step":"Integrate into your workflow.","action":"Add the utility prompt to your `CLAUDE.md` or project documentation. Schedule recurring tasks (e.g., daily WebFetch runs) using cron or CI/CD pipelines.","tip":"For agent creation prompts, define the agent's role, permissions, and termination conditions upfront."},{"step":"Update for new Claude Code versions.","action":"Check the latest system prompts for changes in tool descriptions or utility structures. Modify your prompts to align with new features or deprecations.","tip":"Subscribe to Claude Code's release notes to proactively update utility prompts."}]

Use Cases

Automate the generation of documentation files using CLAUDE.md prompts.

Create custom AI agents tailored to specific tasks using the agent creation prompts.

Perform comprehensive security reviews of code changes to identify vulnerabilities.

Summarize conversations effectively to maintain clarity in team communications.

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/Piebald-AI/claude-code-system-prompts

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

Generate a [UTILITY_PROMPT_NAME] for [PROJECT_NAME] using Claude Code's system prompts. Focus on [SPECIFIC_GOAL] and ensure it follows the structure of existing utility prompts (e.g., CLAUDE.md, compact, statusline). Include [RELEVANT_SECTIONS] such as setup, workflow, and key commands. Reference the latest Claude Code version for compatibility.

Example Output

```markdown
# Utility Prompt: WebFetch for Project Aurora

**Purpose**: Automate web-based data collection for Project Aurora's competitor pricing analysis. This prompt integrates with Claude Code's `WebFetch` tool to scrape and parse pricing data from e-commerce sites while respecting `robots.txt` and rate limits.

## Setup
1. **Prerequisites**: Ensure `claude-code` v0.12.3+ is installed. Add `WebFetch` to your `CLAUDE.md` tools list:
   ```
   tools:
     - name: WebFetch
       description: Fetch and parse web content with rate limiting and error handling
   ```
2. **Environment**: Set `USER_AGENT` in your shell:
   ```bash
   export USER_AGENT="AuroraBot/1.0 (+https://aurora.example.com/bot-info)"
   ```

## Workflow
1. **Target URLs**: Define a list of competitor sites in `./config/competitors.json`:
   ```json
   {
     "sites": [
       "https://competitor-a.com/products",
       "https://competitor-b.io/api/pricing"
     ]
   }
   ```
2. **Execution**: Run the prompt with:
   ```bash
   claude "Fetch competitor pricing data for [PRODUCT_CATEGORY]"
   ```
3. **Output**: Results saved to `./data/pricing/[TIMESTAMP].json` with fields:
   - `product_name`, `price`, `availability`, `last_updated`
   - `source_url`, `status_code`, `fetch_duration_ms`

## Key Commands
- **Validate**: `claude "Check if WebFetch can access https://competitor-a.com/products"`
- **Debug**: Append ` --debug` to any command for verbose logs
- **Rate Limit**: Use `WebFetch --delay 2000` to enforce 2s delays between requests

## Notes
- Respect `robots.txt` by default; override with `--ignore-robots` for internal APIs.
- For JavaScript-heavy sites, add ` --js-render` to enable headless Chrome.
```