How flewis88/webex-ai-bot-via-n8n works
Project type and purpose
• This is an event-driven integration project built on the n8n workflow automation platform.
• Its purpose is to wire up a Cisco Webex Bot that listens for mentions in Webex Spaces, forwards message content to OpenAI via a “Basic LLM Chain” node, and then posts the AI-generated response back into the same Webex space.
File structure
• In this case, there is only a single top-level file (README.md) because the bulk of the “code” lives inside an n8n workflow (a JSON document you import into your n8n instance).
• Key artifacts are:
• Webhook trigger node (captures incoming Webex events)
• Webex by Cisco node (fetches full message content)
• AI Agent / Basic LLM Chain node (invokes OpenAI)
• HTTP Request node (posts replies back to Webex)
• Configuration is managed via n8n credentials (Webex Bot Token, OAuth2 integration for Webex, OpenAI API key).
README for additional insights
• The README walks through:
• Creating a Webex Bot and capturing its bearer token & Bot ID
• Setting up two different Webhooks (test vs. prod) in n8n
• Establishing an OAuth2 integration for the “Webex by Cisco” node
• Calling the OpenAI API via the “Basic LLM Chain” node
• Issuing an HTTP Request to the Webex REST API to send messages
• It shows the sequence in which data flows from Webex → n8n → OpenAI → Webex.
Detailed capture of essential architectural elements
• Two distinct Webex integrations (Webhook creation vs. REST API messaging).
• OAuth2 handshake for the “Webex by Cisco” node.
• Where the OpenAI API key is applied.