A set of command-line utilities that integrate with Todoist and OpenAI to help implement the Getting Things Done (GTD) methodology.
todo-get: Retrieves and displays all your tasks from Todoistnext-action: Analyzes a task and rewrites it as a clear, actionable "next action" following GTD principlesopenai-models: Lists available OpenAI models for use with thenext-actioncommand
-
Clone this repository:
git clone <repository-url> cd gtd -
Install dependencies:
bundle install -
Set up your environment variables:
cp .env.example .env -
Edit the
.envfile with your API keys:- Get your Todoist API token from Todoist settings -> Integrations -> API token
- Get your OpenAI API key from platform.openai.com
Display all your Todoist tasks:
todo-get
Display tasks for a specific project:
todo-get --project "Work"
Rewrite a task as a GTD next action:
echo "Plan vacation" | next-action
This will output something like: "Create a list of 3 possible vacation destinations with available dates and budget estimates"
Process multiple tasks at once (more efficient):
cat tasks.txt | next-action
You can combine these tools:
todo-get | grep "Project" | next-action
Run with verbose output:
echo "Prepare presentation" | next-action --verbose
List all available OpenAI models for your account:
openai-models
The model used for the next-action command can be configured in your .env file by setting the OPENAI_MODEL variable.
These commands work well with standard Unix pipes. Here are some useful examples:
To limit the output to the first 10 lines:
todo-get | head -n 10
Process tasks in OpenAI network call batches of 20 (default 10):
todo-get | head -n 40 | next-action --batch 20
MIT