This collection provides foundational utilities for direct file manipulation and environment management. It enables the reading, writing, and searching of files while automating Python dependency installation via uv pip.
Beyond basic operations, the suite includes specialized tooling for the development of new AI skills. This allows for the creation, optimization, and performance benchmarking of skill descriptions to improve triggering accuracy.
Follow the documented setup, then try a first task.
Installs: Pre-built Memento-S desktop app, including its skill-centric runtime and built-in skill bundle · Memento-S
Platform: Windows
Download the Windows package from the linked Memento-S release page and install it using the platform installer. This is a manual GUI installation; no terminal is required.
Installs: Pre-built Memento-S desktop app, including its skill-centric runtime and built-in skill bundle · Memento-S
Platform: macOS
Download the Mac package from the linked Memento-S release page and install it using the platform installer. This is a manual GUI installation; no terminal is required.
Installs: Complete Memento-Skills source repository and its four built-in skills · Memento-S
Platform: POSIX-compatible shell environment
In a terminal, start in the parent directory where the Memento-Skills repository should be created. Clone the repository, enter it, create and activate its virtual environment, and install the project in editable mode by running this complete block.
git clone https://github.com/Memento-Teams/Memento-Skills.git cd Memento-Skills python -m venv .venv source .venv/bin/activate pip install -e .
On first launch, Memento-S automatically creates `~/memento_s/config.json`. Before starting the app, manually open that file and fill in your model profile. The supplied excerpt does not document a terminal command for editing the configuration.
After filling in the model profile, run the documented diagnostic command in a terminal from the Memento-Skills project directory to check the environment, dependencies, configuration, and API-key availability.
memento doctor # Check environment, dependencies, and config validity
After diagnostics, choose the terminal interface by running this command in the active project environment. Use this or the desktop GUI below.
memento agent # Start an interactive agent session in the terminal
Alternatively, launch the desktop GUI from the active project environment.
memento-gui # Launch the desktop GUI with chat interface
Uses skill-creator
Make sure skill-creator is installed in your agent before trying this task.
Use the skill-creator skill to help me create a skill for [goal]. First narrow down what it should do, then produce an initial draft and a few test prompts. Keep this first pass small and do not run large-scale evaluations.
Uses web-search
Make sure web-search is installed in your agent before trying this task.
Use the web-search skill to fetch [URL] and extract the information relevant to [question]. Return a concise summary and identify the parts of the fetched content that support it. Use URL fetching only; do not perform a web search.
Uses filesystem
Make sure filesystem is installed in your agent before trying this task.
Use the filesystem skill to search files under [directory] for [text or filename pattern]. List the matching paths and briefly summarize the matches without modifying any files.
Claude
OpenClaw