iMessage Bot

Drafts a reply to your most recent iMessage DM or group chat using OpenAI, then previews it before sending. Pick a chat from the tabbed picker (DMs / Group Chats), edit the AI draft, hit Send. If the draft has paragraphs separated by blank lines, each is sent as its own iMessage.

Setup

1. Create a venv

python3 -m venv ~/.venvs/message_bot

2. Install dependencies

~/.venvs/message_bot/bin/pip install -r requirements.txt

3. Add your OpenAI API key

Copy .env.example to .env and fill in your key:

cp .env.example .env

Or, if you'd rather not use a .env file, export it in your shell (or add to ~/.zshrc):

export OPENAI_API_KEY="sk-..."

.env takes precedence over shell exports. Other knobs you can set in .env:

• CUSTOM_DM_PROMPT / CUSTOM_GROUP_PROMPT — override the system prompts (use {name} as a placeholder).
• MULTI_MESSAGE_WEIGHT — 0.0–1.0. 0.0 guarantees a single message, 1.0 guarantees multiple; values in between bias the model smoothly.

Defaults live in messaging.py.

4. Grant Full Disk Access

In System Settings → Privacy & Security → Full Disk Access, add your terminal (and Python if prompted) so the bot can read ~/Library/Messages/chat.db and the AddressBook databases.

Run

~/.venvs/message_bot/bin/python /Users/adityar/Downloads/iMessage-Bot/bot.py

Set AUTO_SEND = True in bot.py to skip the preview window and send immediately (note that this is highly dangerous and discouraged).

Files

• bot.py — entry point / orchestration
• db.py — SQLite reads from chat.db (DMs + group chats) and AddressBook
• gui.py — tabbed picker (DMs / Group Chats) + reply preview
• messaging.py — OpenAI reply generation + AppleScript send (DM and group)
• .env.example — template for OPENAI_API_KEY, CUSTOM_DM_PROMPT, CUSTOM_GROUP_PROMPT, MULTI_MESSAGE_WEIGHT