Featured

Deploy OpenClaw in 60 seconds — 20% off logoDeploy OpenClaw in 60 seconds — 20% off

Launch OpenClaw on Hostinger in about 60 seconds and keep your agent live 24/7. Our referral link gives you 20% off, no coupon code needed.

Launch on Hostinger
Run your Hermes agent on Hostinger, fully managed logoRun your Hermes agent on Hostinger, fully managed

Launch Hermes on Hostinger in one click, fully managed, no VPS knowledge needed. Use code ZACAARON10 for 10% off.

Launch on Hostinger
Crawl and scrape any site into clean data, 10% off logoCrawl and scrape any site into clean data, 10% off

Firecrawl crawls and scrapes any site into clean markdown for your agent. Get 1,000 free credits, and new users get 10% off their first purchase.

Try Firecrawl free
Your own AI agent, running 24/7 with QwikClaw logoYour own AI agent, running 24/7 with QwikClaw

QwikClaw sets up and runs an always-on OpenClaw agent for you. One click, no config files, no server setup.

Deploy now
One API to scrape, enrich, and extract the internet. logoOne API to scrape, enrich, and extract the internet.

Context.dev gives your agents a single API to scrape, enrich, and extract live web data — no proxies, no parsers, no maintenance.

Start building free
SetupClaw: done-for-you OpenClaw for founders & exec teams logoSetupClaw: done-for-you OpenClaw for founders & exec teams

White-glove OpenClaw for founders and exec teams (4–50+ employees): we install, harden, integrate your tools, and maintain it — secured from day one.

Get it set up for you
SEO data APIs for your agent, $1 free credit logoSEO data APIs for your agent, $1 free credit

DataForSEO gives your agent live access to SERP results, keyword data, backlinks, and on-page SEO data through one API. New accounts get a $1 credit, good for up to 20,000 keyword or backlink lookups.

Try DataForSEO free
Reach 47,000+ AI builders

A flat monthly placement in front of developers actively installing AI tools. No lock-in, cancel anytime.

Advertise here

Summary

Agent Instant Messenger for Claude Code. Run multiple agents across local and remote workspaces in a single conversation.

Install to Claude Code

/plugin install aim@aim-marketplace

Run in Claude Code. Add the marketplace first with /plugin marketplace add the-vampiire/agent-instant-messenger if you haven't already.

README.md

AIM — Agent Instant Messenger

![License: MIT](LICENSE)

A Claude Code plugin for multi-agent collaboration. Run multiple Claude Code agents across local and remote workspaces in a single conversation — with automatic push notifications when agents respond.

Install

claude plugin install --from "https://raw.githubusercontent.com/the-vampiire/agent-instant-messenger/master/.claude-plugin/plugin.json"

Quick Start

# Add an agent (creates a session automatically)
/aim:add /path/to/other-project

# Talk to it
@other-project what are you working on?

# The response is delivered automatically — no polling needed

Usage

Add agents

/aim:add /path/to/project              # local workspace
/aim:add /path/to/project:api          # local with alias
/aim:add user@host:/path/to/project    # remote via SSH
/aim:add user@host:/path/to/project:prod-api  # remote with alias

Agent names are derived from the last directory component unless you provide an alias after :.

Talk to agents

@agent-name do something              # message a specific agent
@all what's the status?               # broadcast to all agents (you respond too)

After sending a message, a background sidecar automatically polls for the response and delivers it to your session when it arrives. No manual checking needed.

Manage the session

/aim:status                            # agents, message counts, activity, config
/aim:set @agent permission_mode bypassPermissions  # skip permission prompts
/aim:set @agent timeout_ms 600000      # 10 minute timeout
/aim:set break_timeout_ms 300000       # sidecar polls for 5 min before giving up

How It Works

Terminal A (you)                       Terminal B (agent workspace)
─────────────────                      ──────────────────────────
@backend deploy to staging
  │
  ├─ aim_send fires
  ├─ PostToolUse hook spawns           claude --print runs in
  │  aim-sidecar in background         /path/to/backend workspace
  │                                      │
  │  sidecar polls aim_check ←──────── response written to
  │  every 5 seconds                   shared history.jsonl
  │                                      │
  ├─ sidecar detects new message ◄─────┘
  ├─ delivers to your session
  │
  ▼
@backend: Deployed to staging.
Build #47, all checks passed.

Key concepts

  • Sessions — Created automatically on first /aim:add. Tracks agents, history, and config. Stored at ~/.claude/aim/sessions/<id>/.
  • DEFAULT_AGENT — Your current Claude Code session. You're a participant too — @all messages include you.
  • Fire-and-forget deliveryaim_send returns immediately. The agent runs in the background via claude --print.
  • Sidecar push notifications — A background subagent polls aim_check and delivers responses automatically when they arrive.
  • Session upsert — Agents resume their most recent Claude Code session in the target workspace, preserving context across messages.

MCP Tools

| Tool | Description | |------|-------------| | aim_add | Add an agent to the session (creates session if needed) | | aim_send | Send a message to a specific agent | | aim_broadcast | Send a message to all agents | | aim_set | Configure agent permissions, timeouts, or session settings | | aim_history | View or search conversation history | | aim_sessions | List all sessions with workspace info | | aim_status | Show current session status and config | | aim_check | Poll for new messages (used internally by sidecar) |

Workspace Path Formats

| Format | Example | |--------|---------| | Local | /home/user/projects/backend | | Local + alias | /home/user/projects/backend:api | | SSH | deploy@prod:/app/backend | | SSH + alias | deploy@prod:/app/backend:prod-api |

Configuration

Agent settings (via aim_set)

| Setting | Values | Default | |---------|--------|---------| | permission_mode | default, acceptEdits, bypassPermissions, plan | default | | timeout_ms | Any number >= 1000 | 300000 (5 min) |

Session settings (via aim_set)

| Setting | Values | Default | |---------|--------|---------| | break_timeout_ms | Any number >= 10000 | 120000 (2 min) |

The break timeout controls how long the sidecar polls for a response before giving up. Increase it for agents that take a long time to respond.

Architecture

AIM runs as an MCP server (stdio) with 8 tools. Your Claude Code session is the coordinator. Other agents are spawned via claude --print (local) or over SSH (remote).

~/.claude/aim/
└── sessions/
    └── <uuid>/
        ├── session.json      # agents, config, metadata
        ├── history.jsonl     # public message log
        ├── audit.jsonl       # full log including private messages
        └── check-state.json  # sidecar polling state

SSH agents

For SSH agents, AIM: 1. Probes the remote host for the claude binary path (interactive login shell, then common locations) 2. Finds the most recent Claude Code session in the remote workspace 3. Runs claude --print --resume <session> over SSH with BatchMode=yes

Requires passwordless SSH access (key-based auth) to the remote host.

License

MIT

Related plugins

Browse all →