Personal Ollama Terminal AI

A Zsh-based CLI tool providing direct terminal access to local Ollama models with context management, session persistence, and customizable system behavior.

✨ Features

• 🗣️ Streaming Output: Continuous token-by-token response display from your local Ollama models.
• 🧠 Global Session Management: Maintains a single conversation context across all terminal instances, with automatic token limit handling regardless of which terminal window you use.
• 📝 Global Persistent Memory: System-wide notes file (~/.config/ollama/ai_persistent_notes.txt) accessible to the model across all sessions.
• 🤖 Customizable System Prompt: Define model behavior and parameters via ~/.config/ollama/ai_system_prompt.txt.
• ⚙️ Configurable Settings: Set Ollama model, API endpoint, and context limits in ~/.config/ollama/ai_settings.conf.
• ✍️ Multi-line Input: Support for complex multi-line queries.
• 🛠️ Session Management:
  • Edit notes, system prompts, and settings with ai subcommands (e.g., ai --edit-notes).
  • Reset context (ai --reset).
  • View context info (ai --info context).
• 🚀 Simple Installation: Automated setup via installer script.

✅ Prerequisites

Make sure you have these installed:

1. Ollama: Running with a model pulled (e.g., ollama pull gemma3:4b-it-qat). See ollama.com.
2. jq: JSON processor (e.g., brew install jq or apt-get install jq).
3. curl: Data transfer tool (usually pre-installed).
4. zsh: The Z shell.

🛠️ Installation

1. Clone or Download:

   # If you have git
   git clone <your_repository_url> # Replace <your_repository_url> with the actual URL
   cd personal_ollama_cli
   # If downloaded, navigate to the personal_ollama_cli directory

2. Run the Installer: From the personal_ollama_cli directory:

   ./install.sh

   The installer will:

   • Create config and cache directories (~/.config/ollama, ~/.config/zsh, ~/.cache).
   • Copy ollama_ai.zsh and default configs.
   • Add a source line to ~/.zshrc.

3. Apply Changes: Open a new terminal or run:

   source ~/.zshrc

🚀 How to Use

Interact with your Ollama model using the ai command:

Basic Prompt:

ai Tell me a joke

Multi-line Input: (End with """ on a new line)

ai """
What are the best practices
for writing a good README file?
"""

⚙️ Management Commands

• Settings:

  • ai --show-settings: View current settings.
  • ai --edit-settings: Open settings file in your editor.

• Persistent Notes:

  • ai --view-notes: Show your persistent notes.
  • ai --edit-notes: Edit your notes. Psst! Customize this with your info!

• System Prompt:

  • ai --view-system: Display the system prompt.
  • ai --edit-system: Customize the AI's base instructions.

• Conversation Context:

  • ai --info context: Show context token count and limit.
  • ai --reset: Clear conversation context.
  • ai -r "New prompt": Reset context and send a new prompt.

• Temporary Overrides:

  • ai -m <model_name> "Your prompt": Use a different model for this query (e.g., ai -m gemma3:12b-it-qat "Hi").
  • ai -s "New system prompt" "Your prompt": Use a different system prompt (resets context).

• Help:

  • ai --help or ai -h: Show all commands and options.

🎨 Customization

• ~/.config/ollama/ai_settings.conf: Change default model (AI_OLLAMA_MODEL), API URL (AI_OLLAMA_API_URL), max context tokens (AI_MAX_CONTEXT_TOKENS).
• ~/.config/ollama/ai_persistent_notes.txt: Global information store for the AI across all sessions.
• ~/.config/ollama/ai_system_prompt.txt: Define the AI's personality and default instructions.

🗑️ Uninstalling

1. Remove files from ~/.config/ollama/, ~/.config/zsh/ollama_ai.zsh, and ~/.cache/ollama_ai_context.json.
2. Remove the sourcing line from ~/.zshrc.

Developer

By Soroush Yousefpour

© All rights reserved.

Donate