Imsg logo

Imsg

OrganizationPopular
openclaw
imsg

Use the imsg CLI from OpenClaw agents for iMessage/SMS DMs, groups, replies, reactions, polls, watching, and private-API actions.

Overview

Publisheropenclaw
Repositoryopenclaw
Skill nameimsg
Stars
391K
Forks
82.2K
Bundled files
Instructions only
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • Self-contained

    Everything the model needs lives in the instructions — no extra files to sync.

  • Open source

    Published by openclaw on GitHub. Read the source before you install it.

Installation

Install the Imsg AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
    https://github.com/openclaw/openclaw/tree/main/extensions/imessage/skills/imsg
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/openclaw/openclaw.git /tmp/openclaw
mkdir -p .claude/skills
cp -r /tmp/openclaw/extensions/imessage/skills/imsg .claude/skills/imsg
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Imsg in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Imsg on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Imsg is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

imsg

Use the generic message tool first for actions that its current iMessage schema exposes. Use imsg when the task needs local Messages.app history, target discovery, watching, or an administrative/private-API capability that is not exposed by message.

Do not use this skill for Telegram, Signal, WhatsApp, Discord, Slack, or for replying inside the current OpenClaw conversation when the configured channel already routes the reply.

Agent Flow

  1. Resolve the conversation first.
  2. Choose DM, existing group, or new group.
  3. Prefer an available message action; otherwise pick the lowest-capability imsg command that preserves the requested semantics.
  4. Confirm any send or visible state change unless the user already gave exact recipient, content, and action.
  5. Execute with stable identifiers: prefer --chat-id for normal sends/watch/history and --chat chat GUID for bridge actions.

Never infer a recipient from a casual name alone when several chats or handles could match. Show the matched display name, handle(s), group participants, and message text/action before sending.

Host Requirements

  • macOS 14+ with Messages.app signed in for send/react/bridge actions.
  • Private API mode is strongly encouraged for OpenClaw iMessage. It unlocks replies, precise tapbacks, effects, polls, attachment replies, read/typing actions, and group management. Basic mode is a fallback for reads plus plain text/file send.
  • Full Disk Access for the process context that runs imsg or OpenClaw; reads fail without Messages DB access.
  • Automation permission for Messages.app when using public send.
  • Accessibility permission for the process context that runs public imsg react; it uses System Events UI automation. Bridge tapback uses private API instead.
  • Optional Contacts permission for contact-name resolution.
  • SMS sends require Text Message Forwarding from the user's iPhone to this Mac.
  • Linux reads a copied chat.db only; it cannot send, react, launch Messages.app, mark read, or type.

Resolve Targets

Use --json reads. Output is newline-delimited JSON; use jq -s when jq is available, or consume one object per line directly.

bash
imsg chats --limit 25 --json | jq -s
imsg search --query "dinner" --match contains --json | jq -s
imsg history --chat-id 42 --limit 20 --attachments --json | jq -s
imsg group --chat-id 42 --json

Target rules:

  • DM to a phone/email: use imsg send --to when the user gave an exact handle, or after a single unambiguous chat match.
  • Existing DM thread: imsg send --chat-id <id> is safer than re-resolving a name.
  • Existing group: inspect with imsg group --chat-id <id> --json; send with --chat-id, bridge with the chat GUID from group.
  • New group: use chat-create only when the user explicitly asked to create a group or no existing group matches.
  • Ambiguous group names: confirm participants, not just display name.
  • SMS: use --service sms only when requested or when iMessage fallback is not desired. SMS relay requires Text Message Forwarding.

Do not make jq a hard prerequisite for the skill; it is only a convenient formatter for examples.

Capability Choice

Use imsg standard commands when the current message schema does not cover the operation or when local history is needed:

  • Read/list/search/watch: chats, group, history, search, watch
  • Basic text/file send: send
  • Standard tapback to most recent incoming message in a chat: react

Use the private API bridge for the native iMessage actions OpenClaw users normally expect:

  • Rich replies, text formatting, effects, subjects, multipart sends
  • Native Apple Messages polls and poll votes
  • Tapback by message GUID or tapback removal
  • Edit, unsend, delete, notify anyways
  • Read receipts, typing indicators, bridge event watch
  • Group create/name/photo/member/leave/delete/mark actions
  • Account, whois, nickname checks

For OpenClaw channel setup, check bridge availability early:

bash
imsg status --json

If the host supports bridge actions but Messages is not injected yet, ask before running imsg launch. It kills and relaunches Messages.app to inject the bridge, so treat it as a visible state change:

bash
imsg launch
imsg status --json

If SIP, library validation, private entitlement checks, or missing selectors still block the capability, explain that the requested private-API action is unavailable on this host and offer the closest non-bridge action, if one exists. Do not silently downgrade a threaded reply, effect, subject, poll, or GUID-targeted tapback into a plain send/react.

DM Scenarios

Exact handle, basic send:

bash
imsg send --to "+14155551212" --text "On my way" --service auto

Known DM thread:

bash
imsg send --chat-id 42 --text "On my way"
imsg send --chat-id 42 --file /path/to/photo.jpg

Force channel only when the user asks:

bash
imsg send --to "+14155551212" --text "green bubble" --service sms
imsg send --to "+14155551212" --text "iMessage only" --service imessage --no-sms-fallback

Threaded reply, formatting, effects, or attachment reply:

bash
imsg send-rich --chat 'iMessage;-;+15551234567' \
  --reply-to <message-guid> --text "reply text"
imsg send-rich --chat 'iMessage;-;+15551234567' --text 'hello world' \
  --format '[{"start":0,"length":5,"styles":["bold"]}]'
imsg send-rich --chat 'iMessage;-;+15551234567' --text "boom" --effect impact
imsg send-attachment --chat 'iMessage;-;+15551234567' \
  --reply-to <message-guid> --file /path/to/file.jpg

Formatting ranges are UTF-16 positions. Supported styles include bold, italic, underline, and strikethrough; use --format-file for generated JSON.

Group Scenarios

Inspect before acting:

bash
imsg group --chat-id 42 --json
imsg history --chat-id 42 --limit 20 --json | jq -s

Send to an existing group:

bash
imsg send --chat-id 42 --text "Works for me"

Bridge reply or poll in an existing group:

bash
imsg send-rich --chat 'iMessage;+;chat0000' \
  --reply-to <message-guid> --text "replying in thread"
imsg poll send --chat 'iMessage;+;chat0000' \
  --question "Dinner?" --option "Pizza" --option "Sushi"

Create or mutate groups only on explicit request:

bash
imsg chat-create --addresses '+15551111111,+15552222222' --name 'Crew' --text 'gm'
imsg chat-name --chat 'iMessage;+;chat0000' --name 'Renamed'
imsg chat-photo --chat 'iMessage;+;chat0000' --file /path/to/group.jpg
imsg chat-add-member --chat 'iMessage;+;chat0000' --address +15553333333
imsg chat-remove-member --chat 'iMessage;+;chat0000' --address +15553333333
imsg chat-leave --chat 'iMessage;+;chat0000'
imsg chat-delete --chat 'iMessage;+;chat0000'
imsg chat-mark --chat 'iMessage;+;chat0000' --read

Group mutations are highly visible. Confirm the exact group and participant list before changing membership, name, photo, read state, leaving, or deleting.

Reactions and Replies

Public react is limited: it reacts to the most recent incoming message in the chat.

bash
imsg react --chat-id 42 --reaction like
imsg react --chat-id 42 --reaction love
imsg react --chat-id 42 --reaction dislike
imsg react --chat-id 42 --reaction laugh
imsg react --chat-id 42 --reaction emphasis
imsg react --chat-id 42 --reaction question

For a specific message GUID or removal, use bridge tapback:

bash
imsg tapback --chat 'iMessage;-;+15551234567' --message <message-guid> --kind love
imsg tapback --chat 'iMessage;-;+15551234567' --message <message-guid> --kind love --remove

Use send-rich --reply-to <message-guid> for threaded replies. Confirm the referenced message if the user says "that" or "the previous one".

Polls

Native Apple Messages polls require the bridge. Creation needs at least two --option values. Voting requires one of --option-id, --option-index, or --option.

Messages renders only the options on a poll balloon, so current imsg poll send echoes --question as a best-effort plain caption after the poll. Use --comment to override that caption. Do not retry automatically when only the caption fails: the poll may already be delivered.

bash
imsg poll send --chat 'iMessage;-;+15551234567' \
  --question "Dinner?" --option "Pizza" --option "Sushi" --comment "Vote by 5pm"
imsg poll send --chat 'iMessage;+;chat0000' --reply-to <message-guid> \
  --question "Approve?" --option "Yes" --option "No"
imsg poll vote --chat 'iMessage;+;chat0000' \
  --poll <poll-message-guid> --option-id <option-id>

Find poll IDs and options with:

bash
imsg history --chat-id 42 --limit 20 --json | jq -s '.[] | select(.poll != null) | {guid, poll}'

history and watch backfill a title-less inbound native poll's poll.question from its clean caption row. Poll vote rows are poll events, not tapbacks; watch --reactions is not required to see them.

Watch and Long-Running Agents

For a short one-off wait, use watch:

bash
imsg watch --chat-id 42 --since-rowid 9000 --json
imsg watch --chat-id 42 --attachments --convert-attachments --json
imsg watch --chat-id 42 --reactions --json
imsg watch --chat-id 42 --bb-events --json

--since-rowid is exclusive. Without it, watch starts at the newest row. watch uses filesystem events plus a low-frequency polling fallback, so it can catch up after missed SQLite sidecar events. Poll objects appear without --reactions.

For a daemon or multi-chat integration, use imsg rpc. It speaks JSON-RPC 2.0 over stdin/stdout. Inspect imsg status --json for rpc_methods before using newer bridge or poll methods.

Safety Rules

  • Confirm recipient, chat, and content before every send unless the user's request already contains exact values.
  • Confirm visible state changes: read receipts, typing indicators, edits, unsends, deletes, poll votes, tapbacks, group membership, group name/photo, leaving/deleting chats.
  • Never send to unknown numbers or ambiguous contact-name matches without approval.
  • Confirm attachments exist and are the intended files.
  • Prefer E.164 phone numbers; use --region US or another region only when needed for local formats.
  • Use bridge actions for bridge-only semantics, but confirm visible state changes and destructive actions first.

Frequently asked questions

What does the Imsg AI skill do?

Use the imsg CLI from OpenClaw agents for iMessage/SMS DMs, groups, replies, reactions, polls, watching, and private-API actions.

Why use Imsg on TypingMind?

Because you install it once and use it with any model. Imsg is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Imsg in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/openclaw/openclaw/tree/main/extensions/imessage/skills/imsg. TypingMind reads its SKILL.md and installs it as a skill you can enable per chat.

Which AI models can use Imsg?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Imsg?

As many as you like. As long as a model supports skills, you can use Imsg with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Imsg AI skill free?

It is published on GitHub by openclaw. Check the repository for licensing terms. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇