Outside Agent: a local-first prompt that gets the screen out of the way

Description and use

A CLI tool that uses a local LLM to generate concise outdoor activity plans, reducing screen time and avoiding location tracking.

Problem

Users want a quick outdoor activity plan that minimizes screen time and does not rely on cloud services or location tracking.

Outcome

The tool demonstrates that a local LLM can generate concise, privacy-preserving outdoor activity plans, reducing screen time and avoiding cloud dependencies.

Workflow

• 1. Install Ollama and a local model (default gemma3:4b).
• 2. Run python3 outside_agent.py --area "a nearby public park" --minutes 35 --energy low --interest birds.
• 3. The CLI sends a single request to the local Ollama API at http://127.0.0.1:11434/api/generate with streaming disabled.
• 4. The model returns a Markdown plan containing Mission, What to bring, Screen-off cue, and Adaptation sections.
• 5. The user reads the plan, verifies local conditions, and goes outside without the device.

Evidence

Outside Agent is a small command-line tool that turns a few constraints into one outdoor micro-adventure: a broad area, available time, energy level, and an interest such as birds, plants, or light.

Source

The tool intentionally asks only for a broad area. It does not collect an address, track location, check weather, invent trail conditions, or make claims about opening hours or accessibility.

Source

It sends one request to http://127.0.0.1:11434/api/generate with streaming disabled, then prints the model's Markdown plan.

Source

The project has automated tests for the user-context prompt, its safety contract, and the local Ollama request shape.

Source