Skip to content

CONTRIBUTING.md

Latest commit

 

History

History
126 lines (88 loc) · 3.51 KB

CONTRIBUTING.md

File metadata and controls

126 lines (88 loc) · 3.51 KB

Contributing to Hive Mind (languages: en • zh • hi • ru)

Human-AI Collaboration Guidelines

This project leverages AI-driven development with human oversight. Follow these practices:

Development Workflow

  1. Issue Creation - Humans create issues with clear requirements
  2. AI Processing - Hive Mind analyzes and proposes solutions
  3. Human Review - Code review and architectural decisions
  4. Iterative Refinement - Collaborative improvement cycles

Code Standards

  • TypeScript/JavaScript: Strict typing required
  • File Size: Maximum 1000 lines per file
  • Testing: 100% test coverage for critical paths
  • Documentation: Machine-readable, token-efficient

Version Management with Changesets

This project uses Changesets to manage versions and changelogs. This eliminates merge conflicts that occur when multiple PRs bump the version in package.json.

Adding a Changeset

When you make changes that affect users, add a changeset:

npm run changeset

This will prompt you to:

  1. Select the type of change (patch/minor/major)
  2. Provide a summary of the changes

The changeset will be saved as a markdown file in .changeset/ and should be committed with your PR.

Changeset Guidelines

  • Patch: Bug fixes, documentation updates, internal refactoring
  • Minor: New features, non-breaking enhancements
  • Major: Breaking changes that affect the public API

Example changeset summary:

Add support for automatic fork creation with --auto-fork flag

Release Process

  1. When PRs with changesets are merged to main, the Release workflow automatically creates a "Version Packages" PR
  2. The Version Packages PR updates package.json versions and CHANGELOG.md
  3. When the Version Packages PR is merged, the package is automatically published to NPM

AI Agent Configuration

interface AgentConfig {
  model: 'sonnet' | 'haiku' | 'opus';
  priority: 'low' | 'medium' | 'high' | 'critical';
  specialization?: string[];
}

export const defaultConfig: AgentConfig = {
  model: 'sonnet',
  priority: 'medium',
  specialization: ['code-review', 'issue-solving'],
};

Quality Gates

Before merging, ensure:

  • All tests pass
  • File size limits enforced
  • Type checking passes
  • Human review completed
  • AI consensus achieved (if multi-agent)

Communication Protocols

Human → AI

# Clear, specific instructions
./solve.mjs https://github.com/owner/repo/issues/123 --requirements "Security focus, maintain backward compatibility"

AI → Human

# Status reports with actionable items
echo "🤖 Analysis complete. Requires human decision on breaking changes."

Testing AI Agents

import { testAgent } from './tests/agent-testing.ts';

// Test agent behavior
await testAgent({
  scenario: 'complex-issue-solving',
  expectedOutcome: 'pull-request-created',
  timeout: 300000, // 5 minutes
});

Code Review Process

  1. Automated Review - AI agents perform initial analysis
  2. Cross-Agent Validation - Multiple agents verify solutions
  3. Human Oversight - Final architectural and security review
  4. Consensus Building - Resolve conflicts through discussion

Review Checklist

  • Algorithm correctness verified
  • Security vulnerabilities assessed
  • Performance implications considered
  • Documentation completeness
  • Integration test coverage