JSPM

  • ESM via JSPM
  • ES Module Entrypoint
  • Export Map
  • Keywords
  • License
  • Repository URL
  • TypeScript Types
  • README
  • Created
  • Published
  • Downloads 12
  • Score
    100M100P100Q65016F
  • License MIT

Lightweight CLI for managing AI prompts. Store, organize, and instantly retrieve ChatGPT & Claude prompts. Features tags, variables, sanitization, and optional command execution.

Package Exports

  • cuecli
  • cuecli/src/index.js

This package does not declare an exports field, so the exports above have been automatically detected and optimized by JSPM instead. If any package subpath is missing, it is recommended to post an issue to the original package (cuecli) to support the "exports" field. If that is not possible, create a JSPM override to customize the exports field for this package.

Readme

cueCLI - AI Prompt Management CLI Tool for ChatGPT, Claude & More

npm version License: MIT Node.js Version

Manage, organize, and reuse your AI prompts from the command line

🌐 Website📖 Documentation🐛 Issues📦 NPM

📌 Overview

cueCLI is a command-line tool for managing, organizing, and reusing AI prompts across ChatGPT, Claude, and other assistants. Store frequently-used prompts locally, organize them with tags, and retrieve them instantly. No more copy-pasting from text files or losing prompts in chat history.

Why cueCLI?

Managing prompts across AI chat sessions is messy. Copy-pasting leads to drift, sharing with teams is difficult, and sensitive data needs sanitization. cueCLI solves this by giving you a local prompt library with instant access, smart organization, and built-in safety features. Store once, use everywhere.

Perfect for:

  • Developers using ChatGPT, Claude, or other AI assistants daily
  • Teams needing consistent AI interactions across projects
  • Anyone tired of losing or re-writing prompts
  • Prompt engineers building reusable templates

Key Features

  • Prompt Storage - Save and organize reusable prompts locally
  • Instant Retrieval - Get any prompt to clipboard in under 100ms
  • Tag Organization - Categorize prompts for easy discovery
  • Variable Substitution - Create dynamic templates with placeholders
  • Built-in Sanitization - Automatically detect and remove sensitive data
  • Backup System - Automatic backups of your prompt library
  • Command Execution - Optionally execute prompts as shell commands
  • Cross-platform - Works on macOS, Linux, and Windows

Installation

npm install -g cuecli

From source

git clone https://github.com/cuecli/cueCLI
cd cuecli
npm install
npm link

Quick Start

# Save a frequently-used prompt
cuecli add context --from-file README.md --tags project docs

# Get it back instantly (preview → copy)
cuecli get context

# List all your saved prompts
cuecli list

# Use variables for dynamic prompts
cuecli get api-test --vars API_KEY="..." ENV="staging"

# Execute command prompts safely (optional)
cuecli get deploy-script --execute

Core Commands

cuecli get <name>

Preview-first, then copy-first. In TTY, cueCLI shows a numbered preview (default 10 lines), copies the prompt (or prints it with a clear fallback message), and displays a concise directive summary with a single prompt: “Press Enter to continue; press Esc to clear.” In non‑TTY or machine modes, preview/summary are suppressed for clean output.

Options:

  • --vars KEY=value... - Substitute template variables
  • --stdout - Print to terminal (in non‑TTY suppresses preview/summary)
  • --pipe - Raw output for piping (no preview/summary)
  • --file <path> - Save exact content to file (prints “Saved to ”)
  • --append <path> - Append exact content to file (prints “Appended to ”)
  • --preview / --lines <n> - Show preview first; default 10 lines (TTY)
  • --output <format> - json | markdown | html | base64 | url (machine output)
  • --execute - Only executes truly executable content (shebang or metadata) with confirm: “About to run this prompt via . Run? [y/N]”. Non‑executable (Markdown/prose) is hard‑blocked.
  • --raw / --scan-only - Sanitization controls (see below)

cuecli add <name>

Save a new prompt to your library. After saving, cueCLI shows preview → copy (or fallback print) → directive summary with a single Enter/Esc prompt in TTY.

Options:

  • --from-file <path> - Import from file
  • --from-clipboard - Import from clipboard
  • --desc <text> - Set a short description (shown in list)
  • --tags <tags...> - Add tags for organization

cuecli list

Display all saved prompts with metadata. In TTY, list is numbered and interactive: type an index to retrieve and see the same preview → copy/summary flow. Non‑TTY shows a non‑interactive snapshot.

Options:

  • --tags <tags...> - Filter by tags
  • --json - Output as JSON

cuecli edit <name>

Edit an existing prompt in your default editor. After saving, cueCLI shows preview → copy (or fallback print) → directive summary with a single Enter/Esc prompt in TTY.

Options:

  • --editor <editor> - Use specific editor
  • --desc <text> - Update description without opening editor

Editor notes and troubleshooting:

  • cueCLI uses whatever editor command you configure. Set --editor, $EDITOR, or ~/.cuecli/config.json key "editor".
  • If the editor command is not found, the terminal will show: Editor command not found: <editor> with fixes. Common choices:
    • macOS built‑in: --editor 'open -W -e'
    • POSIX editors: --editor nano or --editor vim
    • VS Code CLI: install the code command, then use --editor 'code -w'.

Tip: When you already have the exact new text, it’s often faster to paste it into your AI assistant and ask it to update the prompt directly (no editor required). Use the terminal for interactive editing or when you prefer your local tools.

cuecli export

Export your prompt library for backup or sharing. For safety, export asks for a one‑line confirmation before writing.

Options:

  • --output <file> - Output file path
  • --raw - Export without sanitization (includes sensitive data)
  • --format <format> - Output format (json, markdown)

Template Variables

Create reusable templates with variables using {{VARIABLE}} or ${VARIABLE} syntax:

Project: {{PROJECT_NAME}}
Environment: ${ENV}
API Endpoint: {{API_URL}}

Then substitute when retrieving:

cuecli get api-template --vars PROJECT_NAME="MyAPI" ENV=staging API_URL="https://api.example.com"

Data Sanitization

Protect sensitive information with built-in sanitization:

#!/usr/bin/env bash
# Scan for sensitive data without modifying
cuecli get my-prompt --scan-only

# Bypass sanitization (raw)
cuecli get my-prompt --raw

# Export with or without sanitization (confirm when prompted)
cuecli export --output shared-prompts.json
cuecli export --raw --output shared-prompts.json

Automatically detects:

  • API keys and tokens
  • Passwords and credentials
  • SSH private keys
  • Credit card numbers
  • Email addresses (optional)
  • URLs with embedded credentials

Configuration

cueCLI stores data in ~/.cuecli/:

~/.cuecli/
├── config.json       # Global configuration
├── prompts.json      # Your prompt library
├── backups/          # Automatic backups
└── templates/        # Prompt templates

Environment Variables

  • CUECLI_CONFIG_DIR - Override default config directory
  • EDITOR - Default editor for prompt editing
  • LOG_LEVEL - Set logging level (ERROR, WARN, INFO, DEBUG)
  • NO_COLOR - Disable colored output

Editors & Troubleshooting

  • Default editor resolution order: --editor flag → ~/.cuecli/config.json$EDITORcode.
  • If your editor isn’t found, cueCLI prints actionable fixes (install/configure an editor CLI, use --editor once, or set $EDITOR / config). No automatic fallback is performed.
  • Recommended defaults:
    • macOS: open -W -e (TextEdit, waits) or code -w (VS Code CLI) or nano/vim.
    • Linux: nano, vim, or vi.
    • Windows: notepad or a configured editor CLI.

Using with AI assistants:

  • Editing via an AI assistant is ideal for deterministic updates (e.g., “replace section X with Y”).
  • Running commands is best done directly in your terminal.

Built-in Universal Prompts

cueCLI comes with 6 universal prompts ready to use:

1. Strict Implementation (strict-implementation)

Ensures precise execution without assumptions or creative interpretations.

cuecli get strict-implementation

2. QA Simulation (qa-simulation)

Comprehensive "1000 users" testing audit for finding bugs and UX issues.

cuecli get qa-simulation

3. Precision Modification (precision-modification)

Surgical updates to existing systems with zero scope creep.

cuecli get precision-modification

4. Change Documentation (change-documentation)

Comprehensive documentation standards for all modifications.

cuecli get change-documentation

5. Single Screen Review (single-screen-review)

Deep dive analysis of a specific screen or component.

cuecli get single-screen-review

6. Progress Tracker (progress-tracker)

Comprehensive task tracking and documentation for multi-step projects.

cuecli get progress-tracker

Usage Guidelines

These prompts work best when:

  • You need precise, accountable work with no surprises
  • You're working with existing systems that shouldn't be disrupted
  • You need comprehensive testing or review
  • You require detailed documentation of changes
  • You want to eliminate assumptions and ensure exact execution

Examples

Save and Use Project Context

# Add your project README as context
cuecli add project --from-file ./README.md --tags documentation

# Use it in your AI chat
cuecli get project

Create Debug Template

# Create a reusable debug template
echo "Debug {{SERVICE}} at {{TIME}} with level {{LEVEL}}" | cuecli add debug --from-clipboard

# Use with different values
cuecli get debug --vars SERVICE=auth TIME="2025-01-04 14:00" LEVEL=verbose

Share Sanitized Prompts with Team

# Export without sensitive data
cuecli export --output team-prompts.json

# Team member imports
cuecli import team-prompts.json

Development

Setup

git clone https://github.com/cuecli/cueCLI
cd cuecli
npm install

Testing

Manual smoke checks (preview‑first, copy‑first, guardrails):

# Seed a prompt
cuecli add strict-implementation --from-file prompts/strict-implementation.md --desc "Strict-mode directive"

# List (interactive in TTY)
cuecli list

# Retrieve (preview → copy/print → summary)
cuecli get strict-implementation

# Non-executable guard (exit 1)
cuecli get --execute strict-implementation; echo $?

Releasing

  • Bump & tag locally (example: patch release):
    • npm version patch -m "chore(release): %s"
    • git push origin main --follow-tags
  • Publish via CI (recommended):
    • Add repo secret NPM_TOKEN (npm Automation token) under GitHub → Settings → Secrets → Actions
    • Push a tag like v1.3.99 (created by npm version), the Release workflow will:
      • Install, lint, test
      • Publish to npm (skips if version already exists)
      • Create a GitHub Release from the tag
  • Manual publish (optional):
    • npm publish --access public
    • Then git tag vX.Y.Z && git push origin vX.Y.Z

Contributing

We welcome contributions! Please see our Contributing Guide for details.

License

MIT © 2025 Alex Kisin

Support

☕ Support cueCLI

Enjoying cueCLI? Consider buying me a coffee to keep this project going!

Buy Me A Coffee

Your support helps maintain and improve cueCLI for everyone. Thank you! 🙏

Acknowledgments

Built with:


Enjoy cueCLI! It's been crafted with care to make your AI workflow smoother.
If it's helped you, star this repo and share it with others!