Agent Skills for Wassenger — automate WhatsApp Business for your business by chatting with Claude, Cursor, Codex, Copilot, or any agent that supports the open Agent Skills format.
A pack of 23 ready-to-install Agent Skills that teach your AI agent how to automate every aspect of WhatsApp Business with Wassenger — onboarding a team, organizing the inbox, sending messages, running campaigns, configuring auto-replies, standing up an AI agent that answers customers (with a human handoff), reporting on inbox analytics, plus seven industry-specific playbooks and migration guides from Twilio and Wati. You describe the outcome in plain English. The agent picks the right Wassenger tools, respects the WhatsApp Business API rules (24-hour window, templates, rate limits), and runs the workflow end-to-end.
Scoped to the official WhatsApp Business API — built for production deployments, not legacy QR-paired devices.
Built on top of wassengerhq/mcp-wassenger, the official Wassenger Model Context Protocol server.
Pick the method that matches your agent — each installs the 23 skills and registers the Wassenger MCP server.
Claude Code (plugin):
/plugin marketplace add wassengerhq/whatsapp-skills
/plugin install wassenger-skills@wassengerhq
Any agent (via the open Agent Skills CLI):
npx skills add wassengerhq/whatsapp-skillsGuided setup — interactive CLI that asks for your API key, validates it, auto-detects your agent, installs the skills, and wires up the MCP server:
npx @wassengerhq/skills init # publishing to npm soonYou'll need a Wassenger API key — get one at app.wassenger.com/developers and set it as WASSENGER_API_KEY (the guided setup does this for you).
| Skill | What it does |
|---|---|
wassenger-setup |
Walks the user through getting an API key, connecting a WhatsApp number, and validating the install. |
wassenger-mcp |
Reference for the Wassenger MCP server — every tool, every parameter, every transport option. |
| Skill | What it does |
|---|---|
wassenger-messaging |
Send text, media, templates, polls, scheduled messages — with WABA template rules baked in. |
wassenger-inbox |
Chat triage: statuses, manual assignment, internal notes, archive/restore, bulk housekeeping. |
wassenger-campaigns |
Bulk broadcasts: segmentation, scheduling, delivery tracking, opt-out. |
wassenger-webhooks |
Subscribe to real-time events with HMAC signature verification and retries. |
wassenger-contacts |
CSV import, CRM sync, segmentation, export, phone validation. |
wassenger-team |
Invite agents, set roles (admin / supervisor / agent), grant per-device access, safe offboarding. |
wassenger-labels |
Full label CRUD with the namespace:value convention, bulk-apply, query, reporting. |
wassenger-routing |
Auto-assignment (round-robin, language-based, skills-based), departments, escalation, fallback. |
wassenger-quick-replies |
Canned-response library — shortcuts (/shipping), media, visibility, bulk import. |
wassenger-auto-replies |
Rule-based automatic responses — welcome, out-of-hours, busy, inactive follow-up. |
wassenger-analytics |
Reports & KPIs from live inbox data — status, agent workload, response/resolution time, volume, backlog, language. |
wassenger-ai-agent |
Stand up an AI (LLM) agent that answers customers 24/7 — inside the 24h window, with a hard human handoff, opt-out, and a kill switch. |
| Skill | For businesses that … |
|---|---|
wassenger-ecommerce |
Run an online store (Shopify, WooCommerce, …) and want WhatsApp for order updates, abandoned cart, post-purchase. |
wassenger-sales-bot |
Need lead qualification, meeting scheduling, follow-up sequences with human handoff. |
wassenger-customer-support |
Operate a multi-agent support inbox with SLAs, escalations, and auto-replies. |
wassenger-real-estate |
Handle property inquiries, viewings, and client nurturing. |
wassenger-restaurants |
Take reservations, send menus, push reminders. |
wassenger-logistics |
Send delivery notifications, tracking updates, proof of delivery. |
wassenger-marketing |
Run segmented broadcasts, growth loops, retention sequences. |
| Skill | For teams that … |
|---|---|
wassenger-migrate-from-twilio |
Are moving a WhatsApp integration off Twilio — API mapping, templates, webhooks, before/after Node & Python. |
wassenger-migrate-from-wati |
Are moving off Wati (wati.io) — session/template sends, named→positional params, broadcasts, webhooks. |
Agent Skills follow a progressive disclosure model. At startup, the agent only loads each skill's name and description (~100 tokens each). When your prompt matches a skill's description, the agent loads the full SKILL.md into context and executes the recipe — calling the Wassenger MCP tools in the documented order.
Example:
👤 "Send a booking reminder to everyone who has a reservation tomorrow at La Tagliatella."
→ Agent activates wassenger-restaurants → reads the reservation-reminder recipe → calls the MCP to fetch tomorrow's reservations from your contacts → sends a pre-approved WABA template to each → reports back with delivery status.
- Node.js ≥ 18
- A Wassenger account with at least one connected WhatsApp number
- An agent that supports the Agent Skills format (Claude Code, Cursor, Codex, Copilot, Gemini CLI, Goose, OpenCode, and 30+ others — see agentskills.io/clients)
PRs welcome — see CONTRIBUTING.md.
MIT © Wassenger