OpenClaw Monade Sales Plugin

@amolsoans/openclaw-monade-sales brings Monade outbound sales operations directly into OpenClaw with a practical, security-first tool surface.

It is built for real calling workflows: launch calls, manage assistants, run campaigns, inspect analytics, and manage user-owned SIP trunks without exposing broad control paths.

Why this plugin exists

Most sales teams do not need another general-purpose integration layer. They need reliable operations inside the assistant they already use:

• Start and control outbound calling workflows quickly
• Keep campaign execution and monitoring in one place
• Access analytics and transcripts with user-level boundaries
• Manage user-scoped trunks without opening global infra controls

This plugin is designed around that operating model.

What you get

• Outbound call actions
• Assistant CRUD and retrieval
• Campaign lifecycle operations (create/update/start/pause/resume/stop/delete)
• Campaign contact management and CSV upload
• Transcript and analytics retrieval
• User-scoped trunk management, including default trunk support

All tools are exposed under the openclaw_* namespace.

Security model

Security is a first-class behavior, not a docs footnote.

• Agent-tool surface only : no CLI commands, no RPC methods, no background services
• Opt-in tool exposure : all tools are registered as optional and require allowlist enablement
• User-scoped operations : workflows are designed for authenticated user ownership boundaries
• No global trunk controls in this plugin : global trunk delete/link remains intentionally out of scope
• Strict base URL validation : scheme + host only, no embedded credentials, paths, query, or hash
• Secret-safe outputs : sensitive fields are redacted from tool result payloads

Installation

openclaw plugins install @amolsoans/openclaw-monade-sales

Recommended user install flow (once published to npm)

# 1. Install the plugin
openclaw plugins install @amolsoans/openclaw-monade-sales

# 2. Configure auth (pick one mode)
# A) user_api_key mode (BetterAuth user API key)
openclaw config set plugins.entries.openclaw-monade-sales.config.authMode user_api_key
openclaw config set plugins.entries.openclaw-monade-sales.config.apiKey ba_YOUR_USER_KEY

# B) service_proxy mode (service token + explicit user UID)
openclaw config set plugins.entries.openclaw-monade-sales.config.authMode service_proxy
openclaw config set plugins.entries.openclaw-monade-sales.config.serviceApiKey svc_YOUR_TOKEN
openclaw config set plugins.entries.openclaw-monade-sales.config.userUid user_abc123

# 3. Restart gateway
openclaw gateway restart

You can also use the OpenClaw Web UI at http://127.0.0.1:18789. The plugin config form appears automatically because uiHints are defined in openclaw.plugin.json.

Configuration

Authentication modes

• authMode: "auto" (default): uses apiKey if present, otherwise serviceApiKey
• authMode: "user_api_key": requires BetterAuth user key in apiKey or MONADE_API_KEY
• authMode: "service_proxy": requires service token (serviceApiKey) and user UID (userUid)

Required fields

• user_api_key mode: apiKey
• service_proxy mode: serviceApiKey and userUid

Optional

• baseUrl: Monade API base host (defaults to https://service.monade.ai)
• defaultAssistantId: fallback assistant for outbound call actions
• timeoutMs: HTTP timeout in milliseconds (1000-60000, default 10000)
• allowInsecureBaseUrl: set true only for trusted internal HTTP environments
• sendUserUidHeader: defaults to true; sends X-User-Uid in service_proxy mode

The plugin can install without keys, but tool calls fail until the selected auth mode has required credentials.

Example config

{
  "plugins": {
    "entries": {
      "openclaw-monade-sales": {
        "enabled": true,
        "config": {
          "authMode": "service_proxy",
          "serviceApiKey": "your-service-token",
          "userUid": "user_abc123",
          "baseUrl": "https://service.monade.ai",
          "timeoutMs": 10000
        }
      }
    }
  }
}

Tool enablement

Because tools are optional by design, explicitly allow what you want to expose:

• allow by individual tool name
• allow by plugin id
• allow by plugin tool group, depending on your OpenClaw tool policy

This gives you tight control over operational blast radius.

Do you need a skill?

Only if you want additional agent instructions for when and how to use plugin tools. The tools come from the plugin itself; SKILL.md adds natural-language guidance and operating context. For most setups, a skill is optional.

Operational notes

• Keep allowInsecureBaseUrl disabled unless you fully trust a private network path.
• Use per-environment API keys and rotate keys on a regular schedule.
• Prefer least-privilege tool allowlists for production agents.

Support

• Issues: repository issue tracker
• Source: this plugin lives in the Monade voice config workspace under MCP/openclaw-monade-sales