Overview

GPTs specialized in knowledge management.

日本語 中文

Demo

Execute code in GAS

Setup Instructions

Run the code in Jupyter

how to set up

Slack Search

Related Articles.

• I want to run a language other than Python on ChatGPT "Code Interpreter"
• Jupyter ✖️ Building a ChatGPT self-hosted "Code Interpreter"

Functions

Below is a list of functions that can be called by the chatbot, along with the necessary environment variables and authentication information.
For details on how to implement your own functions, click here

• jupyter_create_kernel: Create a remote Jupyter Notebook kernel.
  • NGROK_API_KEY.
  • JUPYTER_TOKEN.
• jupyter_execute_code: Execute code in the remote Jupyter Notebook.
  • NGROK_API_KEY
  • JUPYTER_TOKEN.
• jupyter_write_file: Write a file to the remote Jupyter Notebook.
  • NGROK_API_KEY
  • JUPYTER_TOKEN.
• gas_execute_code: Execute code in Google Apps Script.
  • GAS_SPREADSHEET_ID
  • src/data/service_account.json.
• create_notion_page: Create a Notion page with the given title and content.
  • NOTION_SECRET
  • NOTION_DATABASE_ID.
• append_notion_page: Append the content to the Notion page.
  • NOTION_SECRET.
  • NOTION_DATABASE_ID.
• open_youtube_url: retrieve subtitles from Youtube URL.
• open_slack_url: retrieve the message list of a thread from a Slack URL.
  • SLACK_BOT_TOKEN.
• open_notion_url: retrieve the content of the page using NotionAPI.
  • NOTION_SECRET.
• open_url: Open the given URL and retrieve its content. Used for web page scraping and API responses.
• open_slack_canvas_url: retrieve content from the slack canvas url.
  • SLACK_BOT_TOKEN.
  • SLACK_USER_TOKEN.
• google_search: search for information from the entire web using Google's custom search engine. Returns search results based on specific keywords.
  • GOOGLE_CSE_ID.
  • GOOGLE_API_KEY.
• arxiv_search: searches for articles using arXive.
• location_search: Finds the latitude and longitude of an address based on keywords.
• trend_search: Performs a Google trend search.
• youtube_search: Search Youtube videos.
• google_drive_search: Search for file names from Google Drive. Returns search results based on keywords.
  • src/data/service_account.json.
• slack_search: Search the message history of a specific channel or user using Slack's API.
  • SLACK_BOT_TOKEN
  • SLACK_USER_TOKEN.
• notion_search: Use Notion's API to search pages and databases within Notion.
  • NOTION_SECRET.
• github_search: Search for Issues, Pull Requests, and code in GitHub and return results.
  • GITHUB_TOKEN.
• get_github_discussion: allows you to retrieve a list of discussions from a Github Discussion URL.
  • GITHUB_TOKEN.
• fact_checker: use Google's Fact Check Tools API to investigate the veracity of a particular claim or news item.
  • GOOGLE_API_KEY
• intelx_search: uses Intelligence X's API to search for information from various data sources (web pages, forums, documents, etc.).
  • INTELX_API_KEY
• ingest_memory: memorize important content of a conversation.
  • PINECONE_API_KEY.
  • OPENAI_API_KEY.
• semantic_search_memory: search the semantic content of the conversation.
  • PINECONE_API_KEY
  • OPENAI_API_KEY

Deploy

cp env.sample .envrc
direnv allow
npm install -g serverless
serverless plugin install -n serverless-api-gateway-throttling
serverless plugin install -n serverless-prune-plugin
aws ecr-public get-login-password --region us-east-1 | docker login --username AWS --password-stdin public.ecr.aws
sls deploy

Jupyter notebook

git clone https://github.com/tatsuiman/jupyter-ngrok-worker
cd jupyter-ngrok-worker
cp env.sample .env
vim .env
vim docker/nginx/nginx.conf
docker-compose up -d --build

Create GPTs

GPTs can be set up by reading here to set them up.