中文版 | English Version
Welcome to MoFA's GSoC 2026 page.
- GSoC 2026 Organization Page: https://summerofcode.withgoogle.com/programs/2026/organizations/mofa-org
- Project Ideas: ideas-list.md (中文版)
- Discord: https://discord.gg/hKJZzDMMm9
- Organization Contact: dev@mofa.ai
All contributors will be featured on mofa.ai Developer Hall of Fame
Can I start working on ideas now? Yes. You do not need to wait for the official GSoC application period. Start exploring the codebase, pick up open tasks, and discuss your approach with mentors. Early contributions help both you and us.
Core framework (Rust):
git clone https://github.com/mofa-org/mofa
cd mofa
cargo buildStudio application (Rust + Makepad):
git clone https://github.com/mofa-org/mofa-studio
cd mofa-studio
./run.shThe legacy Python/Dora-based core (
pip install mofa-core) is still available but is no longer the focus of active development. GSoC projects target the Rust codebase.
- Read
docs/architecture.mdin the mofa repository - Study
examples/in mofa-rs:monitoring_dashboard,multi_agent_coordination,react_agent - Build and run Studio locally
Before applying:
- Open an issue in the relevant repository to discuss your approach
- Email mentors directly (see project listings)
- Organization contact: dev@mofa.ai
You DON'T need:
- To be a Rust expert (we're learning too)
- Prior experience with Makepad or Dora
- Perfect knowledge of AI/LLM internals
You DO need:
- Willingness to learn Rust and systems programming
- Stable time commitment throughout GSoC
- Comfort with uncertainty (we're exploring alongside you)
- Ability to communicate progress and blockers
Bonus points:
- Systems programming experience (any language)
- UI/UX development background
- AI/ML project experience
- Prior open source contributions
Don't aim for perfection. Show us you:
- Understand the problem space
- Have a plausible path forward
- Can break the work into milestones
- Are prepared to adapt as requirements evolve
Include time for:
- Learning Rust/Makepad (Weeks 1-2)
- Understanding the codebase (Week 3)
- Core implementation (Weeks 4-9)
- Testing, documentation, demo (Weeks 10-12)
Tell us:
- Why MoFA interests you
- What you hope to learn
MoFA is a project in transition. We are rewriting our core in Rust, exploring new UI frameworks, and evolving our architecture as we learn. We are looking for people who want to learn and build together with us, not applicants who already know everything.
What we actually care about:
- Can we communicate? If you get stuck, will you tell us? If we explain something poorly, will you ask? That matters more than your resume.
- Will you show up? Consistent effort over the summer beats a brilliant start followed by silence.
- Can you learn on the fly? Nobody on our team was a Rust expert when we started either. We care about your willingness to figure things out, not what you already know.
Issue assignment rules:
- Issues are assigned to the first person who comments expressing interest, by default.
- If you want an issue created or already assigned by someone else, comment with your reason — maintainers will reassign if there is a compelling case.
- If an assignee has no visible progress and no communication for 7 days, maintainers may reassign the issue.
What happens if multiple people want the same GSoC idea?
Multiple people may submit proposals for the same idea. We will talk to each of you and try to find a way — maybe scoping the project differently, maybe suggesting an alternative. We would rather work something out than turn people away.
Do I need a pre-GSoC contribution to apply?
It helps, but it is not a hard requirement. A small PR or a thoughtful issue comment shows us you can navigate the codebase and communicate — that is the point, not gatekeeping.
We strongly encourage you to make at least one contribution before submitting your proposal. This helps us assess your ability and helps you understand the codebase.
-
Comment on the issue expressing your interest. Briefly describe your approach.
-
Wait for assignment. The issue is assigned to the first person who comments, by default. Don't start coding until assigned. If the issue is already assigned and you believe you have a compelling reason, comment to explain — maintainers will reassign if appropriate.
-
Fork and create a branch:
git checkout -b your-branch-name
-
Open a Draft PR early. This shows your progress and lets mentors give feedback.
-
Test your changes locally before marking as "Ready for Review".
-
Respond to feedback. Code review is a conversation, not a verdict.
- Discord: https://discord.gg/hKJZzDMMm9 — introduce yourself in #general, GSoC discussion in #gsoc
- Email: dev@mofa.ai — for longer discussions or private matters
- GitHub Issues — technical questions, bug reports, feature requests
See proposal-template.md for a structured template to guide your application.
MoFA embraces AI tooling and encourages contributors to use it to improve productivity and quality. However, contributors must avoid irresponsible and unethical practices. You are ultimately responsible for all output generated with the help of AI — review it, understand it, and make sure it meets the project's standards.
For detailed guidance, please refer to Google's official document: Guidance for GSoC Contributors using AI tooling in GSoC 2026.
MoFA (Modular Framework for Agents) is an open-source framework for building AI agents.
Our journey started with a Python/Dora-based prototype that taught us what works (composition, controllable code) and what needs improvement (the difficulty of orchestrating and debugging atop an unoptimized substrate). We're now building MoFA Studio — a desktop application built with Rust and Makepad for creating and running AI applications, with OminiX-MLX for on-device ML inference on Apple Silicon — while simultaneously exploring mofa-rs, a next-generation core that may one day serve as an alternative foundation.
Website: mofa.ai
Key Repositories:
- Core runtime: github.com/mofa-org/mofa
- Studio application: github.com/mofa-org/mofa-studio
- Makepad UI components: makepad-chart, makepad-d3, makepad-flow, makepad-element
MoFA is an open-source project under the MIT/Apache-2.0 license.