Slack - supercog-ai/community GitHub Wiki

1. RAG backend with knowledge from: internet, docs, Slack messages, email
2. Document/image analysis
3. Use in private or in shared channels ("project manager")
4. People research including LinkedIn data
5. Customizable instructions and tools per channel

Install page: https://app.supercog.ai/slack/

• Ask the agent what it can do, or ask what tools are available
• Use "{info}" system command to see current status (what tools the agent is using)
• Use "{knowledge}" system command to see current docs are in the RAG
• To connect Google Calendar, or Google Docs, or JIRA, etc.. you use the web interface. Open the "Home" tab for the Slack app and find the Admin Login button:

---

The intro video on that page is a useful walkthrough also.

To add files to RAG, open the "Doc Index" called "personal".

To use Supercog in a shared channel just invite it into the channel (/invite @supercog). This will create a separate knowledge index just for that Slack channel (the index will show on the Connections web page under the name of the channel).

In a public channel, currently you have to @ mention the agent to talk to it, but it will reply in a thread and respond to every message in the thread (unless you use 🤫).

More things to try:

• Drag an image into Slack and ask Supercog to read it for you
• Now connect JIRA. Drag a screenshot of an angry customer email into Slack and say "Please create a JIRA task from this email".
• Try out Text-to-speech tool
• Ask for news from google news
• Ask the agent to read a web page and summarize it (you can also add web site crawls to RAG)
• Extract info, save to CSV or PDF
• Connect a database for text-to-SQL and automatic charts