jay-code/src/cli/init/directory-structure.ts

Streamlined AI CLI orchestration engine with mathematical rigor and enterprise-grade reliability

// init/directory-structure.ts - Directory structure creation
export async function createDirectoryStructure(): Promise<void> {
  const fs = await import('fs/promises');
  const path = await import('path');

  // Define directory structure
  const directories = [
    '.claude',
    '.claude/commands',
    '.claude/commands/swarm',
    '.claude/commands/sparc',
    '.claude/logs',
    '.claude/memory',
    '.claude/configs',
    'memory',
    'memory/agents',
    'memory/sessions',
    'coordination',
    'coordination/memory_bank',
    'coordination/subtasks',
    'coordination/orchestration',
    'reports',
  ];

  // Create directories
  for (const dir of directories) {
    try {
      await fs.mkdir(dir, { recursive: true });
      console.log(`  ✅ Created ${dir}/ directory`);
    } catch (error: unknown) {
      if ((error as any).code !== 'EEXIST') {
        throw error;
      }
    }
  }

  // Create README files for key directories
  const readmeFiles = {
    'memory/agents/README.md': createAgentsReadme(),
    'memory/sessions/README.md': createSessionsReadme(),
    'coordination/README.md': createCoordinationReadme(),
    'reports/README.md': createReportsReadme(),
  };

  for (const [filePath, content] of Object.entries(readmeFiles)) {
    await fs.writeFile(filePath, content);
    console.log(`  ✅ Created ${filePath}`);
  }

  // Create initial persistence database
  const initialData = {
    agents: [],
    tasks: [],
    swarms: [],
    lastUpdated: Date.now(),
    version: '1.0.71',
  };

  await fs.writeFile('memory/jay-code-data.json', JSON.stringify(initialData, null, 2));
  console.log('  ✅ Created memory/jay-code-data.json (persistence database)');
}

function createAgentsReadme(): string {
  return `# Agents Directory

This directory stores persistent information about AI agents created and managed by Jay-Code.

## Structure
- Each agent gets its own JSON file named by agent ID
- Agent files contain configuration, state, and memory
- Shared agent data is stored in agent-registry.json

## Usage
Agents are automatically managed by the Jay-Code orchestration system. You can:
- View agent status with \`jay-code agent list\`
- Create new agents with \`jay-code agent spawn <type>\`
- Configure agents with \`jay-code agent configure <id>\`

## Files
- \`agent-registry.json\`: Central agent registry
- \`agent-<id>.json\`: Individual agent data files
- \`templates/\`: Agent configuration templates
`;
}

function createSessionsReadme(): string {
  return `# Sessions Directory

This directory stores information about Jay-Code orchestration sessions.

## Structure
- Each session gets its own subdirectory
- Session data includes tasks, coordination state, and results
- Session logs are automatically rotated

## Usage
Sessions are managed automatically during orchestration:
- Start sessions with \`jay-code start\`
- Monitor sessions with \`jay-code status\`
- Review session history with \`jay-code session list\`

## Files
- \`session-<id>/\`: Individual session directories
- \`active-sessions.json\`: Currently active sessions
- \`session-history.json\`: Historical session data
`;
}

function createCoordinationReadme(): string {
  return `# Coordination Directory

This directory manages task coordination and orchestration data.

## Subdirectories
- \`memory_bank/\`: Shared memory for agent coordination
- \`subtasks/\`: Task breakdown and assignment data
- \`orchestration/\`: High-level orchestration patterns

## Usage
Coordination data is used for:
- Multi-agent task distribution
- Progress tracking and monitoring
- Resource allocation and balancing
- Error recovery and failover

Access coordination data through the Jay-Code API or CLI commands.
`;
}

function createReportsReadme(): string {
  return `# Reports Directory

This directory stores output reports from swarm operations and orchestration tasks.

## Structure
- Swarm reports are stored by operation ID
- Reports include execution logs, results, and metrics
- Multiple output formats supported (JSON, SQLite, CSV, HTML)

## Usage
Reports are generated automatically by swarm operations:
- View recent reports with \`jay-code swarm list\`
- Check specific reports with \`jay-code swarm status <id>\`
- Export reports in different formats using \`--output\` flags

## File Types
- \`*.json\`: Structured operation data
- \`*.sqlite\`: Database format for complex queries
- \`*.csv\`: Tabular data for analysis
- \`*.html\`: Human-readable reports
`;
}