OpenViSTAI · Agents · Blockchain · Crypto
Submit a project
Menu
Agent Frameworks

Clickclickclick

Clickclickclick is an MIT-licensed Python framework for building agents that perform prompted tasks through Android and computer interfaces using local or remote language models.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Clickclickclick — OpenViST project profile in Agent Frameworks.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Use it when you want to build or explore a model-driven agent that carries out prompted tasks through Android or computer interfaces and can configure a Python environment, model access and, for Android, ADB.

Who it is for

  • Developers building agents to automate Android interactions.
  • Developers exploring model-driven computer-use workflows such as browser tasks or drafting email.

Documented uses

  • Translate a natural-language instruction into steps for an Android or computer interface.
  • Use a finder model to identify interface elements and an executor to act on them.
  • Choose separate planner and finder models, including documented Ollama, Gemini and OpenAI options.
Before you choose

Requirements & limits

  • The framework distinguishes planning, interface-element finding and action execution.
  • The README lists local Ollama models alongside Gemini and GPT-4o options.
  • Android execution requires ADB installed on the machine running the code.
  • Model performance varies by role; the README warns that qwen3.5:4b is unreliable as a finder.

Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.

Source-based practical guide

Install and configure Clickclickclick

Prepare the Python project and configure planner and finder models before running tasks.

Before you start

  • Python 3
  • ADB installed on the machine for Android use
  • Model choices and any API keys required by those models

Steps

  1. Clone the repository and enter its project directory.

  2. Create and activate a Python virtual environment.

  3. Install dependencies with pip install -r requirements.txt.

  4. Run python main.py setup and choose planner and finder models, providing required API keys.

Expected outcome

The project dependencies are installed and the planner and finder have been configured for subsequent task use.

Guide limitations

  • These setup steps do not establish that a particular task will execute successfully.
  • The README describes the code as highly experimental.
Documentation 1 ↗
Source-based practical guide

Submit a task through the command line

Run a natural-language task prompt with selected platform and planner and finder models.

Before you start

  • Clickclickclick installed and configured
  • ADB on the local machine when using Android
  • Configured model access and any required API keys

Steps

  1. Use the documented run command with a task prompt, such as python main.py run "Open Reddit".

  2. Set the platform with --platform=android or --platform=osx as needed.

  3. Choose planner and finder models with --planner-model and --finder-model.

  4. Optionally set --image-quality to a value from 1 to 100.

Expected outcome

The CLI receives the prompt and selected platform and model settings for task execution.

Guide limitations

  • The README describes the framework as highly experimental.
  • It reports that qwen3.5:4b is slow for basic planning and unreliable for UI element finding.
Documentation 1 ↗
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
16 October 2026
Original website title
instavm/clickclickclick · Official homepage title · retained in its source language
How we review sourcesSuggest a correction