neuralmesh / apimesh

Serves as the template to enable llms in any gihub project
GNU General Public License v3.0
0 stars 2 forks source link

prompts #14

Open m-c-frank opened 7 months ago

m-c-frank commented 7 months ago

this serves as the promptlibrary where i post my interesting conversations with chatgpt that could be turned into prompt templates

m-c-frank commented 7 months ago

okay one good way to refactor nicely is:

current file:

filecontent

refactor out this part and update the current file

perhaps give more context by just using the issue title adn description and comments

m-c-frank commented 7 months ago

https://chat.openai.com/share/23ac29c2-371e-45b7-8f16-6550cee7b8a0

m-c-frank commented 7 months ago

be clear and concise and neutral and professional. you dont have much time. increase the information/token density of your response and dont waste my time.

m-c-frank commented 7 months ago

but put everything even the file creations into one bash script. i.e. it will be just a readme with one script that does everythign including the creating of dockerfile and server.py. but all as oneliners

m-c-frank commented 7 months ago

reapplying label comment

m-c-frank commented 7 months ago

reapplying label comment

m-c-frank commented 7 months ago

reapplying label comment

m-c-frank commented 7 months ago

reapplying label comment

m-c-frank commented 7 months ago

use heredocs whenever possible

m-c-frank commented 7 months ago

Based on the principles outlined in the ChatDev paper, we're looking to leverage GitHub as our primary frontend for development workflows. Our objective is to use Langchain expression language chains that are activated by GitHub issue events to automate and route tasks within our company. Create a simple workflow to manage incoming issues using basic code. The main goal is to run a script via the GitHub CLI triggered by issue events that can also pass through the event ID to provide context. The workflow described in the 'taskrouter' flow provided needs to be simple yet efficient. Here's what we need:

Configure a GitHub Action workflow that responds to issue-related events such as opened, labeled, milestoned, assigned, and unassigned, as well as issue comment events. Set up the job to run on the latest Ubuntu environment with appropriate permissions. Establish an environment variable setup that securely includes the OpenAI API key and the GitHub Token. Design a script that works as a 'Task Router', making use of open-source resources when possible, and ensuring this script is hosted publicly, possibly at this URL: https://github.com/neuralmesh/apimesh/blob/main/taskrouter.sh. The script should facilitate Langchain expression language commands, parse events via the GitHub CLI, and automate responses, ensuring it's able to pass and utilize the event ID for context. Integrate error handling and logging mechanisms to monitor and troubleshoot the operations. Please outline a development timeline, including the crucial milestones from inception to the first operational prototype, as well as the resources required at each step."

In absence of specifics from the ChatDev paper, your AI-enhanced executive assistant could use this prompt to create a detailed action plan for your company, setting up the necessary infrastructure and protocols for these GitHub-triggered automation tasks using the GitHub CLI and Langchain expression language. If there are specific points from the ChatDev paper that are crucial to this task, please extract those points and share them, and they can then be included in the prompt refinement