Quick Start Guide

Get up and running with HelseCLI in 5 minutes.

Quick Start Guide

This guide describes the procedures for initial deployment and execution of HelseCLI workflows.

Launching HelseCLI

To initiate the HelseCLI interactive environment, execute the following command in your terminal:

python helsecli.py

Upon successful execution, the HelseCLI home screen will be displayed:

   ╦ ╦╔═╗╦  ╔═╗╔═╗  ╔═╗╦  ╦
   ╠═╣║╣ ║  ╚═╗║╣   ║  ║  ║
   ╩ ╩╚═╝╩═╝╚═╝╚═╝  ╚═╝╩═╝╩

   AI-powered coding assistant • v2.0.0

  ╭─────────────────────────────────────────────────────────╮
  │               Welcome to Helse                          │
  │         Your AI-powered coding assistant                │
  ╰─────────────────────────────────────────────────────────╯

  Select operational mode:

    [1] Code Mode     - Autonomous programming and research
    [2] Document Mode - Technical and scientific document synthesis
    [3] Chat Mode     - General AI interaction

  ─────────────────────────────────────────────────────────

    [H] Help & FAQ     - System documentation
    [Q] Quick Start    - Operational guide access
    [X] Exit           - Terminate session

  Enter selection (1/2/3/H/Q/X):

Operational Modes

Code Mode

Select 1 to enter Code Mode. This environment is optimized for technical development and research.

Capabilities:

  • Program synthesis and refactoring
  • Multi-file contextual analysis
  • Sandboxed code execution
  • Automated web research
  • Version control integration

Input Examples:

> Generate a Python utility for CSV parsing and data visualization.

> Analyze the current authentication module for potential security vulnerabilities.

> Identify all pending 'TODO' items within the project scope.

Document Mode

Select 2 to enter Document Mode for structured writing and report generation.

Capabilities:

  • Academic and scientific paper synthesis
  • Technical documentation generation
  • Business and research report compilation
  • Multi-format export (Markdown, DOCX, PDF)

Configuration Workflow:

  1. Define document classification (e.g., "Technical Specification").
  2. Select citation standards (e.g., "APA 7th Edition").
  3. Specify target word count.
  4. Select export format.
  5. Configure web research parameters.
  6. Define research objective or topic.

Chat Mode

Select 3 for general inquiries. This mode is a lightweight interface for direct LLM interaction without local tool access.

Input Examples:

> Explain the implementation of asynchronous patterns in Python.

> Summarize industry best practices for RESTful API architecture.

Initial Task Validation

Follow these steps to validate the installation and functionality in Code Mode:

1. Initialize Code Mode

python helsecli.py code

2. Execute Instruction

Provide a specific task for the assistant:

> Create a Python script titled 'validation.py' that outputs a system status message.

3. Execution Audit

The system will log the following operations:

I'll initialize the validation script.

[Action: write_file]
  Path: validation.py
  Payload: print("HelseCLI System Validated")

Success: File generated at validation.py

Core System Commands

The following commands are available across all operational modes:

CommandFunction
!exitTerminate current mode
!homeReturn to primary interface
!helpAccess command reference
!mModel selection interface
!model <name>Explicit model switch
!keysAccess keyboard shortcut reference

Code Mode Specific Commands

CommandFunction
!aToggle Autonomous Agent Mode
!stepsDisplay action audit trail
!planEnable Planning Phase (Plan-Before-Execute)
!buildEnable Direct Execution Mode
!verifyToggle automated code verification

System Hardware/Keyboard Shortcuts

ShortcutFunction
Ctrl+DToggle execution pause
Ctrl+SSave state or resume
Ctrl+CTerminate current process

Best Practices for Optimal Output

  1. Precision: Provide detailed context. Avoid ambiguous instructions.
  2. Autonomous Execution: Utilize Agent Mode (!a) for complex, multi-stage workflows.
  3. Audit Trails: Regularly inspect executed actions using !steps to ensure accuracy.
  4. Model Selection:
    • GPT-4o: Recommended for complex programming.
    • Claude 3.5 Sonnet: Optimized for long-form synthesis and analysis.
    • Gemini 2.0 Flash: Optimized for low-latency tasks.

Support Resources

  • Access the internal Help system via H from the home screen.
  • Reference the Tools Directory for technical details on available modules.
  • Consult the Troubleshooting Guide for error resolution.

Previous: ← Installation | Next: Configuration →


Previous: ← Installation | Next: Configuration →