These are (most of) my TypingMind Plugins, I've decided to make them OpenSource for anyone to use who wants them.
All plugins are designed to have a AWS Lambda Python function part and a simple wrapper for calling the Lambda function on the TypingMind side.
Installation
- Create an AWS account (you can do this for free, you get a lot of AWS Lambda function calls completly free each month)
- Create a lambda function
- Give it a name that is easily identifiable
- Set the runtime to Python (preferably 3.10 or 3.11 both should work)
- Under advanced settings click Enable function URL
- Set Auth type to None (We'll have own custom auth in most plugins, however be aware that this makes your functions publicly accessible in theory)
- Enable "Configure cross-origin resource sharing (CORS)"
- Configure lambda function
- Go to configuration/General -> Set timeout according to each function (you can probably just set this to around 1m)
- Go to configuration/Function URL and click on edit
- Set allow Origin to * (or you can set it to the typingmind domain)
- Add x-api-key, access-control-allow-origin, and content-type to "Expose headers" and "Allow headers"
- Under Allow methods add *
- Enable Allow credentials
- Go to configuration/Environment variables -> Set your API Keys and set the "SERVICE_API_KEY" to what you want your API key to be
- Copy the python code to your functions lambda_function.py
- Look under the plugin header below which packages are needed for the plugin
- Open a Linux Terminal (Under WSL or native)
- Create a folder
mkdir python
- Run
python3 -m pip install package_1 package_2 --target ./
- Zip the folder
cd .. && zip -r my-lambda-layer.zip python
- Create a new Layer on AWS, select the python version and upload your ZIP file
- Copy your function URL
- Create a new Plugin on Typingmind
- Copy the respective plugin-name.json to the OpenAI function spec
- Copy the respective plugin-name.js to the Code Implementation and replace with your function URL and API Key
- And you're done!
A simple YouTube transcript getter, useful for Video Summarization etc.
Installation
- Packages: youtube_transcript_api
An in-depth LLM search engine with multiple modes.
Chat example: https://chatgpt.com/share/98ab9702-369f-42e4-b982-23345dc7b4d7
Installation
- Packages: bs4, requests, re, openai, boto3
- Set your "OPEN_ROUTER" API key
- Set your "BRAVE_SEARCH_TOKEN" API key
- Set your "SCRAPING_ROBOT_TOKEN" API key
A simple Wolfram|Alpha llm api wrapper
Installation
- Packages: requests
- Set your "WOLFRAM_APP_ID" API key
Wrapper for multiple ImageGen models.
Installation
- Packages: replicate, requests, openai
- Set your "OPENAI_API_KEY"
- Set your "REPLICATE_API_KEY"
Python Interpreter (using a notebook like output) with Ipython.
Installation
- Packages: IPython + any packages you may aditionally want for a code interpreter