[DISCUSS] Call for GSoC 2026 Project Ideas

[Baoyuan <ba...@apache.org>]

Hi Community,

Google Summer of Code (GSoC) 2026 is around the corner. GSoC is a fantastic
way to welcome new contributors and complete features that have been on our
wishlist for a while. We are now calling for project ideas from everyone in
the community.

Please reply to this email or list your ideas in this thread. If you are a
committer and willing to mentor a specific task, please let us know!

Re: [DISCUSS] Call for GSoC 2026 Project Ideas

[DEEP SHEKHAR SINGH / <24...@iiitu.ac.in>]

Subject: Re: Call for GSoC 2026 Project Ideas

Hi Community,

I would like to propose a project idea for GSoC 2026 regarding the APISIX
Dashboard.

Project Title: Support JSON Schema to Form UI Generation

Description:
Currently, APISIX has over 80+ plugins, and adding UI support for new ones
requires manually creating and maintaining separate React components. This
manual process is not only time-consuming but also prone to inconsistencies
across such a large ecosystem. This project aims to solve this by building
a system that automatically generates Form UIs directly from existing JSON
Schemas.

Why this fits GSoC:
This is a substantial undertaking that involves not just building a complex
UI adapter, but also ensuring it handles the edge cases of 80+ different
plugin schemas (nested arrays, conditional logic, custom validation).
Successfully standardizing this across the entire dashboard is a
large-scale infrastructure improvement perfect for the GSoC timeline.

Goals:
1.  Develop a Schema Adapter: Create a robust component that renders
dynamic forms based on standard JSON Schema definitions.
2.  Advanced Schema Support: Implement support for complex features like
`oneOf`, `dependencies`, and `type: "array"` to cover the requirements of
sophisticated plugins.
3.  Validation System: Integrate `ajv` to bridge the gap between schema
validation and UI error handling.
4.  Mass Migration Strategy: Validate the system by migrating complex
plugins (e.g., `proxy-rewrite`) and establish a pattern for migrating the
remaining 80+ plugins.

Current Status:
I have actively researched this in Issue #2986 and developed a Proof of
Concept ( PR #3274 ) which demonstrates the feasibility of handling
conditional rendering (`oneOf`) using the existing tech stack.

I believe this project will significantly improve the maintainability of
the Dashboard and future-proof the plugin system.

Best regards,
Deep
Github: DSingh0304