2.5k
GitHub Stars
3
Bundled Files
2 months ago
Catalog Refreshed
4 months ago
First Indexed
Readme & install
Copy the install command, review bundled files from the catalogue, and read any extended description pulled from the listing source.
Installation
Preview and clipboard use veilstrat where the catalogue uses aiagentskills.
npx veilstrat add skill openclaw/skills --skill autoresponder- _meta.json279 B
- README.md13.6 KB
- SKILL.md14.0 KB
Overview
This skill monitors iMessage and SMS conversations on macOS and automatically replies to selected contacts using AI-generated responses shaped by per-contact prompts and recent conversation context. It includes rate limiting, time windows, keyword triggers, testing mode, and Telegram-based management for on-the-fly configuration. Use it to automate replies while keeping control over tone, frequency, and privacy.
How this skill works
A background watcher listens for incoming messages via the imsg CLI, checks the watch list for configured contacts, fetches recent conversation history, and generates a reply using the configured AI prompt. Rate limiting, daily caps, time windows, and keyword rules are applied before sending with imsg send. All actions, state, and history are logged and can be managed via Telegram commands or CLI utilities.
When to use it
- Auto-reply to recurring messages when you can't respond immediately (e.g., driving, in meetings).
- Maintain a consistent voice across responses to family, coworkers, or clients using per-contact prompts.
- Filter replies to only urgent messages using keyword triggers and time windows.
- Test AI-generated responses before sending with the preview/test mode.
- Manage responders remotely via Telegram slash commands or natural language.
Best practices
- Add only contacts who expect automated replies or where automation is appropriate.
- Create concise, specific prompts for each contact to get consistent voice and length.
- Use rate limits and daily caps to avoid spamming and preserve trust.
- Regularly review /autorespond_history to audit what the skill sent.
- Run the watcher in background (screen/tmux or nohup) and monitor logs for errors.
- Keep API keys and Full Disk Access correctly configured and test imsg locally before enabling automation.
Example use cases
- Automatically send a short, warm acknowledgment to a parent with a custom prompt during work hours.
- Reply with a brief, sarcastic quip to a close friend using a tailored prompt and short daily cap.
- Only respond to messages containing keywords like 'urgent' or 'help' while ignoring casual chats.
- Preview AI replies for a new prompt with /autorespond_test before enabling live replies.
- Remotely add or disable contacts from the watch list using Telegram natural language commands.
FAQ
macOS with Messages.app signed in, imsg CLI installed, Full Disk Access for your terminal, and a configured AI API key.
How do I prevent the skill from spamming a contact?
Set per-contact minMinutesBetweenReplies, enable daily caps, and configure time windows and keyword triggers.