ollama-agentic 0.1.8

A beautiful, agentic CLI for Ollama — run local LLMs with auto tool-calling, memory, and more

ollama-agentic

A beautiful, agentic terminal interface for Ollama — run local LLMs with auto tool-calling, long-term memory, iterative code debugging, and more.

Install

pip install ollama-agentic
ollama-cli

Ollama is installed and started automatically if not already present.

---

Features

• ⚡ Auto mode — model autonomously calls tools to complete tasks (/auto)
• 🔁 Iterative debug loop — /run file.py auto-fixes errors until code passes
• 📋 Plan executor — /plan <goal> breaks goals into typed steps and executes them
• 🧠 Long-term memory — /remember stores facts that persist across sessions
• 📦 Auto-installs Ollama — detects if Ollama is missing and installs it for you
• 🚀 Auto-starts Ollama — spins up ollama serve automatically if not running
• ⬇️ Arrow-key model picker — /install lets you browse and download 25+ models
• 🔧 Agent tools — /shell, /file, /fetch, /ls inject real context into chats
• 💾 Conversation saving — /save and /load persist chats as JSON
• 🎭 Personas — save and load system prompt presets
• 🆚 Compare mode — run the same prompt through two models side by side

---

Usage

ollama-cli                       # start chatting
ollama-cli --model qwen2.5:7b    # start with a specific model
ollama-cli --auto                # start in autonomous agent mode
ollama-cli --compare             # compare two models side by side

---

Commands

Chat & Navigation

Command	Description
/cls	Clear screen (keep context)
/clear	Clear conversation and screen
Ctrl+L	Clear screen
/retry	Regenerate last response
/tokens	Toggle token count display

Models

Command	Description
/model	Switch active model (arrow-key picker)
/current	Show currently active model
/install	Browse & install models from catalogue
/models	List all installed models
/compare	Compare two models side by side

Agentic

Command	Description
/auto	Toggle autonomous tool-calling mode
/plan <goal>	Break a goal into steps and execute
/run <file.py>	Run code, auto-fix errors in a loop

Memory

Command	Description
/remember <fact>	Store a fact in long-term memory
/memories	List all stored memories
/forget <id>	Delete a memory by ID

Context Injection

Command	Description
/file <path>	Load a file into context
/shell <cmd>	Run a shell command, inject output
/fetch <url>	Fetch a webpage into context
/ls <path>	Inject a directory listing
/context	View or clear active injections

Conversations & Personas

Command	Description
/save <n>	Save conversation
/load <n>	Load conversation
/list	List saved conversations
/system <prompt>	Set a system prompt
/persona <n>	Load a saved persona
/personas	List saved personas
/save-persona <n>	Save current system prompt as persona

---

Agent Mode

Toggle with /auto or launch with --auto. In auto mode the model can call tools, read results, and loop until the task is done — no manual /file or /shell needed.

⚡ you › look at main.py and find any bugs
⚡ you › write a web scraper for hacker news and run it
⚡ you › set up a basic Flask app in this folder

---

Config & Data

All config and data is stored in your home directory:

Path	Description
~/.ollama_cli_config.json	Settings (model, auto mode, etc)
~/.ollama_cli_history	Input history
~/.ollama_cli_memory.json	Long-term memories
~/.ollama_cli_saves/	Saved conversations
~/.ollama_cli_personas/	Saved personas

---

Requirements

• Python 3.10+
• macOS, Linux, or Windows
• Ollama (handled automatically on first run)

---

Roadmap

• MCP server — expose tools to Claude Code, Cursor, and other agents
• Repo-aware context — auto-index codebase on launch from a project folder
• Git tools — /diff, /commit, /log
• API key integrations — Claude, OpenAI, Gemini, Groq as model backends
• Symbol search across codebase

---

Contributing

PRs and issues welcome at github.com/Akhil123454321/ollama-cli. Keep changes focused and include tests where appropriate.

License

MIT — see LICENSE