Closed seriybeliy11 closed 7 months ago
How the application works:
🚀 GitHub API Endpoints for Project Profiling:
Contributors:
Endpoint: https://api.github.com/repos/ton-society/grants-and-bounties/contributors
Description: Fetches information about contributors to the specified project.
User Information:
Endpoint: https://api.github.com/users/{name}
Description: Retrieves essential user information, including name, biography, public repositories, follower and following counts.
User Repositories:
Endpoint: https://api.github.com/users/{name}/repos
Description: Retrieves a list of user repositories, including names and descriptions.
Repository Languages:
Endpoint: https://api.github.com/repos/{name}/{repo_name}/languages
Description: Fetches information about programming languages used in a specific repository.
These endpoints are critical to the profiling engine because they provide comprehensive information about participants, user details, user repositories, and programming languages associated with specific projects. The combination of these data sources forms the basis for effective profiling of projects and recruitment processes (for example, for applications)
As a result, we have a simple front-end in which each contributor will be displayed in the format:
This is useful for our project because it is important to see everyone’s strengths and, if necessary, involve people in specific tasks in order to increase the reliability and efficiency of solutions. This will simplify management and allow participants to get to know each other better. We bet $700, because for the processing and parsing itself, a minimalistic interface will be made in the React Material Kit.
@delovoyhomie implementation period: 1 week
@delovoyhomie
As the community and project audience grows, this tool can be useful for assessing the skills of new and existing participants
I have concerns about the practicality and value of adding another interface for a single GPT prompt, especially considering the $700 cost. Additionally, I find that GPT is not adequately equipped to assess a developer's performance effectively, particularly given that the prompt relies heavily on GitHub profile metrics like follower count and public repositories, which are not robust indicators of a developer's true skills and experience.
Summary
Implement an AI-based system to compile short summaries for Github contributors of TON Society
Bounty Context
Why it's Important? Having an automated system for summarizing the contributions made by each developer can save a lot of manual work and help the team quickly understand each person's role. This would furthermore, encourage involvement by making contributions more visible and understandable.
Problem showcase Currently, there's no easy way to get a concise summary of what contributions a particular developer has made to the TON Society. Going through each commit or change manually is time-consuming and inefficient.
Potential Solution Propose to build an AI-based system that processes all the contributions made by a developer and automatically generates a short summary. This system would use a powerful language model like GPT-3 to make sense of the change logs and compile the summaries. It would also bundle this with a straightforward web interface allowing anyone in the team to quickly get summaries for any contributor.
Defenition of Done
References
Please refer to similar projects such as: (1) how GitHub autogenerates summaries for repositories using machine learning algorithms, and (2) the implementation of GPT-3 for natural language processing tasks.
REWARD
🪄 Created with the help of TON Bounties Creator Bot.