Send results and images to your local WhatsApp Desktop account via a surgical Python bridge. Use when you need to output data to WhatsApp contacts or groups without UI interference.
Install
npx skillscat add bobbyflow/whatsapp-messenger-skill Install via the SkillsCat registry.
SKILL.md
WhatsApp Messenger
This skill allows AI agents to send text output and images directly to your local WhatsApp Desktop client using a hardware-resilient Python bridge.
🌟 Key Features
- ⚡ Atomic Delivery: Uses high-speed clipboard injection to bypass physical keyboard interference.
- 🛠️ Self-Healing: Automatically launches WhatsApp if it's closed or hidden in the system tray.
- 🛡️ Absolute Zero Aggression: Forces WhatsApp to front and locks hardware input (Requires Admin).
- 🔄 Follow-the-Focus: Automatically detects and re-binds if a chat is in a standalone window.
Tools
send_to_whatsapp
Sends a message or image to a specific WhatsApp contact.
Arguments:
contact: The name of the contact or group (e.g., "Bobby").message(optional): The text to send.image(optional): The absolute local path to an image file.send(optional): Set to true to actually transmit (default: false / Halt mode).
Command:
python "%USERPROFILE%\.gemini\skills\whatsapp-messenger\scripts\whatsapp_bridge.py" "<contact>" [--message "<message>"] [--image "<path>"] [--send]
Workflow
- Identify the target contact and content.
- Use the
send_to_whatsappcommand to deliver. - Default (HALT Mode): The script MUST stay in Halt Mode (pastes but doesn't send).
- Bypass Rule: The AI Agent is STRICTLY FORBIDDEN from using the
--sendflag unless the user explicitly includes the phrase "BYPASS HALT" in their request.
Constraints
- WhatsApp Desktop Required: The Windows Desktop application must be installed.
- Phone Connection: Your phone must be connected to ensure the desktop client is in an active state.
Bobby Choi Sovereign | Opal Architect