My Daily Log

Log what you do every day. It's a good habit!

Setup

The first time you use the extension, you will be prompted to set up a directory to store your logs. You can also change this later in the Raycast preferences. Your logs are plain JSON files (one per day), so you can put the folder in iCloud Drive or Dropbox to sync it between machines.

Commands

• New Log: log what you just did. Type the title as an argument to log it instantly, or open the form to pick another date/time and log something you forgot.
• My Daily Log: the logs of a day, with actions to edit, delete, copy the day as Markdown, and move between days (⌘ [ / ⌘ ]). The optional argument accepts t (today), y (yesterday), a number of days ago (3), a weekday (fri) or a date (2022-12-31).
• Logged Days of Month: every day you logged something, grouped by month.
• Search Logs: search everything you have ever logged.
• Log Reminder: a menu bar item that reminds you to log when you haven't logged anything for a while. You can configure the interval, your working hours, weekends and notifications, and snooze it from the menu.
• Day Summary, Summary of a Week, Summary of a Month and Day Standup Speech: AI generated summaries, weekly reports and standup speeches.

AI features with your own (local) model

The AI commands don't use Raycast AI. They work with any server that exposes an OpenAI-compatible API, so your logs can stay on your machine.

Ollama (default)

1. Install Ollama and start it.
2. Download a model, e.g. ollama pull llama3.2.
3. That's it: the default preferences point to http://localhost:11434/v1 and the llama3.2 model. Change the AI Model preference to use any other model you have (ollama list).

Other providers

Set the AI Server URL, AI Model and, if needed, AI API Key preferences:

Provider	AI Server URL
Ollama	http://localhost:11434/v1
LM Studio	http://localhost:1234/v1
llama.cpp server	http://localhost:8080/v1
OpenAI	https://api.openai.com/v1
OpenRouter	https://openrouter.ai/api/v1

Thinking and extra model parameters

Reasoning models (e.g. qwen3, deepseek-r1, gpt-oss) can be slow because they "think" before answering. Set AI Thinking to Disabled to turn it off (the equivalent of ollama run qwen3 --think=false).

Use Extra Model Parameters to send any other parameter supported by your server with every request, either as flags or as JSON:

--think=false --temperature=0.2 --top_p=0.9

{ "temperature": 0.2, "seed": 42, "max_tokens": 800 }

think is translated to the reasoning_effort parameter that OpenAI-compatible APIs (including Ollama's) expect.

Use the Extra AI Instructions preference to customise every answer, e.g. Answer in Spanish or Use at most 5 bullet points.

Upgrading from older versions

Older versions named the log files after the UTC date, so logs written in the evening or right after midnight could show up on the wrong day. The first time you open this version, your logs are moved to the files of their local date. A backup of the previous files is kept in a hidden .backup-… folder inside your logs folder.