Getting Started¶

Prerequisites¶

Go 1.21+
For voice: Twilio account + ngrok account
For chat: Discord/Telegram bot token (optional)

Installation¶

Build from Source¶

git clone https://github.com/plexusone/agentcomms.git
cd agentcomms
go mod tidy
go build -o agentcomms ./cmd/agentcomms

Configuration¶

AgentComms uses a single JSON configuration file for all settings.

1. Generate Configuration¶

# Full config with voice settings
./agentcomms config init

# Minimal config (chat only, no voice)
./agentcomms config init --minimal

This creates ~/.agentcomms/config.json.

2. Set Environment Variables for Secrets¶

# Discord token
export DISCORD_TOKEN=your_discord_bot_token

# Telegram token (if using)
export TELEGRAM_BOT_TOKEN=your_telegram_bot_token

# Voice providers (if using voice)
export TWILIO_ACCOUNT_SID=ACxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
export TWILIO_AUTH_TOKEN=your_auth_token
export ELEVENLABS_API_KEY=your_elevenlabs_key
export DEEPGRAM_API_KEY=your_deepgram_key
export NGROK_AUTHTOKEN=your_ngrok_authtoken

3. Edit Configuration¶

Edit ~/.agentcomms/config.json:

{
  "version": "1",
  "agents": [
    {
      "id": "claude",
      "type": "tmux",
      "tmux_session": "claude-code",
      "tmux_pane": "0"
    }
  ],
  "chat": {
    "discord": {
      "enabled": true,
      "token": "${DISCORD_TOKEN}",
      "guild_id": "YOUR_GUILD_ID"
    },
    "channels": [
      {
        "channel_id": "discord:YOUR_CHANNEL_ID",
        "agent_id": "claude"
      }
    ]
  }
}

4. Validate Configuration¶

./agentcomms config validate

Running the MCP Server (OUTBOUND)¶

The MCP server enables AI agents to call humans and send chat messages.

./agentcomms serve

Output:

Starting agentcomms MCP server...
Voice providers: tts=elevenlabs stt=deepgram
Chat providers: [discord telegram]
MCP server ready
  Local:  http://localhost:3333/mcp
  Public: https://abc123.ngrok.io/mcp

Running the Daemon (INBOUND)¶

The daemon enables humans to send messages to AI agents.

./agentcomms daemon

Output:

INFO starting daemon data_dir=/Users/you/.agentcomms
INFO database initialized path=/Users/you/.agentcomms/data.db
INFO router initialized
INFO registered agent agent_id=claude type=tmux
INFO daemon started

Claude Code Integration¶

Option 1: MCP Configuration¶

Add to ~/.claude/settings.json:

{
  "mcpServers": {
    "agentcomms": {
      "command": "/path/to/agentcomms",
      "env": {
        "TWILIO_ACCOUNT_SID": "ACxxx",
        "TWILIO_AUTH_TOKEN": "xxx",
        "NGROK_AUTHTOKEN": "xxx",
        "DISCORD_TOKEN": "xxx",
        "ELEVENLABS_API_KEY": "xxx",
        "DEEPGRAM_API_KEY": "xxx",
        "AGENTCOMMS_AGENT_ID": "claude"
      }
    }
  }
}

Option 2: Generate Plugin Files¶

go run ./cmd/generate-plugin claude .

This creates:

.claude-plugin/plugin.json - Plugin manifest
skills/phone-input/SKILL.md - Voice calling skill
skills/chat-messaging/SKILL.md - Chat messaging skill

Testing the Setup¶

Test Outbound (Agent → Human)¶

In Claude Code, the AI can use:

initiate_call: Call your phone
send_message: Send a Discord/Telegram message

Test Inbound (Human → Agent)¶

# Check daemon is running
./agentcomms status

# Send a test message
./agentcomms send claude "Can you also add unit tests?"

# In Claude Code, AI can check for messages:
# Uses check_messages MCP tool

Test Bidirectional¶

Start daemon in tmux session claude-code
Run Claude Code in that tmux session
Send message via Discord
AI uses check_messages to see your message
AI responds via send_message