Skip to content

CONTRIBUTING.md

Latest commit

 

History

History
83 lines (52 loc) · 3.26 KB

CONTRIBUTING.md

File metadata and controls

83 lines (52 loc) · 3.26 KB
Raw

Contributing

Thank you for your interest in the Cybernetic Your Agent project. This project aims to make the ideas from Qian Xuesen's Engineering Cybernetics easily deployable and beneficial for anyone using modern AI agents.

Contributions of any form are welcome.

Contribution Areas

Here are ways you can contribute to this project:

Supplementing Practical Cases

Deploy this architecture on different agent platforms and share your experience:

  • Platform-specific adaptation methods (tool calling syntax, memory mechanism differences, etc.)
  • Problems encountered in practice and their solutions
  • Effectiveness feedback after running for a while (correction repeition rate, feedforward hit rate, etc.)

Theory Extensions

Introduce new theoretical frameworks to extend the L1 meta-cognitive layer:

  • For example: Bayesian inference, Shannon information theory, TRIZ, constraint theory, etc.
  • Requirement: Explain how the new theory works with the existing four concepts, and provide executable L1 extension entries

Translation

Translate the tutorial into other languages. Please first open an Issue to declare the language you are translating, to avoid duplicate work.

Documentation Improvements

  • Fix typos, grammar errors, unclear expressions
  • Supplement diagrams (architecture diagrams, data flow diagrams)
  • Improve example files

Code and Scripts

  • Provide automatic deployment scripts for specific agent platforms
  • Provide tools for memory/skill state export

Submission Guidelines

Issue

  • Before submitting bugs, improvement suggestions, or new ideas, please search existing issues first.
  • Clearly describe the problem or suggestion and include relevant context.

Pull Request

  1. Fork this repository.
  2. Create a separate branch for new features or fixes (e.g., feat/add-claude-mcp-guide).
  3. Keep commit messages concise and clear (e.g., docs: fix integral control threshold explanation in chapter 3).
  4. Link the corresponding issue (if any) by using Closes #issue-number in the PR description.
  5. If adding new content, please follow the tone and structure of existing documents:
    • Concise, direct, engineering-oriented
    • If there's theory, there must be implementation; if there's design, there must be instructions
  6. Ensure independent readability: Each document file should be self-contained and not rely on conversational context.

Content Principles

This tutorial adheres to the following principles. Please ensure compliance when contributing:

  • Executable over theoretically complete: Content provided must be something readers can immediately operate.
  • Architecture over prompt volume: We add structure, not prompt entries.
  • Open-ended endings are allowed: We don't require ultimate solutions to every problem, but we must give troubleshooting directions.

Code of Conduct

  • Discussions should focus on the tutorial itself, maintaining respect and constructiveness.
  • Rational discussion of different opinions is welcome; personal attacks are not.
  • Maintainers reserve the right to close issues/PRs that deviate from the topic or do not follow the guidelines.

License

By contributing, you agree to place your contribution under this project's MIT License.