jason810496 commented on issue #62500: URL: https://github.com/apache/airflow/issues/62500#issuecomment-4105895760
Hi all, Thank you everyone who joined the discussion. I would like to make this thread focus on discussing questions and clarifying project goals. It would be better not to overwhelm this issue discussion thread by adding a new reply here for every new feature or commit to your PoC, as this might cause the real questions for discussion to be pushed away or overlooked, thanks. Since I will still go through all the proposals submitted via the official GSoC platform, you can add a link to your PoC branch in your final proposal and I will definitely take a look at it. Here are some of my opinions, suggestions, and resources that you can use to verify the proposal yourself, as there are a lot of "AI hallucinations" not only about what "agent skills" are, but also about the project goals of most of the PoCs so far. This is really an open-ended project, and it is worthwhile to spend time on the design itself, the trade-offs, and all the possible alternative approaches over the implementation itself. - What are "Agent Skills" (https://agentskills.io/home) ? And what _might_ be the problem of `AGENTS.md` or MCP? - Does your approach really solve the core pain points and align with the project goal? - How will you verify that your approach really works? - How does your approach make the contributor docs the source of truth for agent skills? (Or, how does your approach ensure consistency between the contributor docs and the agent skills?) - How will your approach affect the long-term maintainability for the project maintainers? -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
