Show HN: Pluqqy – Terminal based context management tool for AI coding

Build and manage minimal viable context for your AI coding assistant from your terminal.

Pluqqy lets you create reusable components (contexts, prompts, and rules) and combine them into pipelines. When you set a pipeline, it generates a PLUQQY.md (configurable) file that contains all your composed instructions.

In Claude Code or other AI coding tools, just reference @PLUQQY.md instead of copying and pasting prompts. Need different instructions? Set a different pipeline - the file updates automatically, but you keep referencing the same @PLUQQY.md.

This approach keeps your context minimal and focused - only including what's relevant for the current task. Both you and the AI work from the same single source of truth, eliminating confusion about which instructions are active while preserving context space.

This will build and install pluqqy to $GOPATH/bin or $HOME/go/bin (if GOPATH is not set).

To update to the latest version:

Or manually:

This creates the following structure:

The examples command installs pre-built components and pipelines that demonstrate useful approaches and common patterns. All examples are prefixed with example- to distinguish them from your own components.

Pluqqy provides a comprehensive set of CLI commands for managing pipelines and components without the TUI. All commands support multiple output formats (--output json|yaml) and global flags (--quiet, --yes, --no-color).

All commands support these global flags:

When components share the same filename across different types:

When creating a new component (press n in the components pane):

1. Select Type: Choose between Context, Prompt, or Rules
2. Enter Name: Provide a descriptive name for your component
3. Edit Content: Opens the enhanced editor with all features available:
   • Multi-line editing with line numbers
   • File references (@ to insert)
   • Undo support (^z)
   • External editor integration (^x)
   • Save (^s) or cancel (Esc)

The enhanced editor provides a simple editing experience for both creating new components and editing existing ones.

Keyboard Shortcuts:

Note: This editor is intentionally minimal. For writing new components from scratch or making substantial edits, press ^x to use your configured external editor (vim, VS Code, etc.)

Pluqqy uses the same keyboard shortcuts across all operating systems (macOS, Linux, Windows):

• All shortcuts use Ctrl (^) as the primary modifier key
• ^s for save, ^d for delete, ^x for external edit, etc.
• Single letter shortcuts (like n for new, e for edit) work everywhere

Linux Users:

• If you experience issues with Ctrl shortcuts (e.g., ^s pausing output), disable terminal flow control:
  stty -ixon # Add to your .bashrc or .zshrc
• For clipboard support (y key), install xclip:
  sudo apt install xclip # Ubuntu/Debian sudo dnf install xclip # Fedora/RHEL
  See Linux Clipboard Setup for details.

Pluqqy can generate interactive HTML-based Mermaid diagrams to visualize your pipeline structure:

1. Select any pipeline in the main list view
2. Press M to generate an interactive diagram
3. The diagram opens automatically in your default browser

The visualizer creates a flowchart showing:

• Pipeline name and total component count at the top
• Components organized into subgraphs by type
• Color coding: Contexts (blue), Prompts (green), Rules (red)

Generated diagrams are saved in .pluqqy/tmp/diagrams/ with timestamped filenames for easy reference.

Pluqqy includes a tagging system to help organize and find your components and pipelines:

The built-in search engine supports powerful queries with keyboard shortcuts:

Search Shortcuts:

Note: In Pipeline Builder, ^t cycles through component types only (skips pipelines).

The set command generates a PLUQQY.md file in your project root. This naming convention is intentional:

Yes, I know, the world already has plenty of AGENTS.md, CLAUDE.md, and README.md files. Adding PLUQQY.md might feel like contributing more noise. The good news is you don’t have to call it PLUQQY.md. You can name it whatever you like (AGENTS.md, WIZARD.md, BANANA.md, PLUQQYSUX.md), and you can customize exactly where Pluqqy generates it.

The file contains sections in your configured order:

• CONTEXT - Combined context components
• PROMPTS - Combined prompt components
• RULES - Combined rules components

Pluqqy includes a built-in library of example components and pipelines to help you get started quickly. These examples demonstrate useful patterns and approaches I've found helpful when working with AI coding assistants (they are not best practices).

All example components include:

• Placeholders: Replace {{PLACEHOLDER}} values with your project-specific information
• Clear documentation: Each component explains its purpose and usage
• Practical patterns: Approaches that have worked well in real projects
• Reusability: Designed to work across different projects

1. Replace placeholders: Update {{PROJECT_NAME}}, {{LANGUAGE}}, etc.
2. Modify content: Adapt examples to your specific needs
3. Remove prefix: Rename files to remove the example- prefix when ready
4. Create variations: Use examples as templates for project-specific components

1. Initialize project: pluqqy init
2. Launch TUI: pluqqy
3. Press n to create new pipeline
4. Name it "my-assistant"
5. Add components using the builder
6. Press ^s to save
7. Press S to set the pipeline
8. Check PLUQQY.md for the composed output

Pluqqy includes a built-in settings editor accessible from the TUI. Press s from the main list view to customize:

• Output Settings

  • Default filename for generated output (default: PLUQQY.md)
  • Export path for pipeline output files (default: ./ - your project root)
  • Output path for pipeline-generated files (default: .pluqqy/tmp/)

• Formatting Options

  • Toggle section headings in output
  • Reorder sections using J/K keys
  • Edit section types and headings

Changes take effect immediately upon saving with ^s.

Pluqqy uses your system's $EDITOR environment variable to determine which external editor to use. Set it in your shell configuration (.bashrc, .zshrc, etc.):

Important notes:

• GUI editors (VS Code, Cursor, Windsurf, Zed, Sublime) require a "wait" flag to block until the file is closed
• Different editors use different wait flags: --wait for VS Code/Cursor/Windsurf, -w for Zed/Sublime
• After setting $EDITOR, reload your shell config: source ~/.zshrc (or ~/.bashrc)
• Test your setting by running: echo $EDITOR

How External Editing Works:

1. Press ^x in the enhanced editor to open your external editor
2. A status message appears: "Editing in external editor - save your changes and close the editor window/tab to return here and continue"
3. Make your changes and save the file in the external editor
4. Close the editor window/tab to return to Pluqqy:
   • Terminal editors (vim, nano): Exit normally and you'll return immediately
   • GUI editors (VS Code, Zed, etc.): Close the specific tab or window to return
5. Your changes will be automatically reloaded in Pluqqy

• Go 1.19 or higher (for building)
• Terminal with UTF-8 support

MIT