Espanso HubEspanso Hub

LLM ask AI

llm-ask-ai

by Bernhard Enders

An Espanso package that enables users to quickly send prompts to a local (e.g. Ollama or LM Studio) or remote LLM API calls (OpenAI standard) and insert the AI-generated response directly into any text field.

LLMAIartificial intelligencepromptEspansoOllamaOpenAI
espanso install llm-ask-ai

llm-ask-ai

An Espanso package that enables users to quickly send prompts to a local (e.g. Ollama or LM Studio) or remote LLM API calls (OpenAI standard) and insert the AI-generated response directly into any text field.

Requirements

  • Espanso installed and running
  • Python 3.9+ installed and available on your system path
  • Required Python packages: openai and python-dotenv (see requirements.txt)
  • Access to a local LLM (such as Ollama or LM Studio) or a remote OpenAI compatible API (in this case, you will need to provide your API key in the .env file)

Configuration

Edit the .env environment file inside the package directory () to set the API_KEY, BASE_URLand MODEL. Example:

API_KEY=ollama
BASE_URL=http://localhost:11434/v1
MODEL=llama3.2

NOTE: don't forget to pull the Ollama model first. In the case above, just issue the command:

ollama pull llama3.2

Usage

  • Type the Espanso trigger for this package (:ask:ai) in any text field.
  • Enter your desired AI prompt when asked.
  • The AI-generated response will be inserted automatically (this can take several seconds, so choose a small and low latency LLM model).

Example

Type:

:ask:ai

Then enter:

Summarize the following text: ...

Press the "Submit" button or "CTRL + Enter" to send the request. The response from your configured LLM will appear in place.

Troubleshooting

  • Make sure python (not python3) executable is available globally on your system's PATH environment variable and that the required packages are installed.

  • Certain Linux distributions, such as Debian, use only a python3 executable and don't include a python executable or symlink. In this case, you may need to install the python-is-python3 package (recommended), create a symbolic link, or run a command like this:

    sudo update-alternatives --install /usr/bin/python python /usr/bin/python3 10

  • Make sure you have the necessary Python packages installed globally, either by running pip install openai python-dotenv or by using a command like sudo apt install python3-openai python3-dotenv.

  • Check that your BASE_URL endpoint, MODEL and API_KEY are correctly set in the .env file (located in the Espanso config directory: %CONFIG%/match/packages/llm-ask-ai/.env).

  • After sending the request, make sure the cursor doesn’t lose focus and remains blinking at the correct insertion point. If it doesn’t, just click to place it right after the trigger string like this :ask:ai|

  • Review Espanso logs for errors.

License

MIT

Author

Bernhard Enders

Links