@appkit/llamacpp-cli 1.11.0 → 1.12.1

package/src/commands/router/start.ts

@@ -1,60 +0,0 @@
-import chalk from 'chalk';
-import { routerManager } from '../../lib/router-manager';
-import { stateManager } from '../../lib/state-manager';
-
-export async function routerStartCommand(): Promise<void> {
-  console.log(chalk.blue('▶️  Starting router...'));
-
-  try {
-    // Initialize
-    await routerManager.initialize();
-    await stateManager.initialize();
-
-    // Check if router already exists
-    const existingConfig = await routerManager.loadConfig();
-    if (existingConfig && existingConfig.status === 'running') {
-      console.log(chalk.yellow(`⚠️  Router is already running on port ${existingConfig.port}`));
-      return;
-    }
-
-    // Start router
-    await routerManager.start();
-
-    // Get updated config
-    const config = await routerManager.loadConfig();
-    if (!config) {
-      throw new Error('Failed to load router configuration after start');
-    }
-
-    // Get running servers to show available models
-    const servers = await stateManager.getAllServers();
-    const runningServers = servers.filter(s => s.status === 'running');
-
-    // Display success
-    console.log();
-    console.log(chalk.green('✅ Router started successfully!'));
-    console.log();
-    console.log(chalk.dim(`Endpoint: http://${config.host}:${config.port}`));
-    console.log(chalk.dim(`Available models: ${runningServers.length}`));
-
-    if (runningServers.length > 0) {
-      console.log();
-      console.log(chalk.dim('Models:'));
-      runningServers.forEach(server => {
-        console.log(chalk.dim(`  • ${server.modelName} (port ${server.port})`));
-      });
-    } else {
-      console.log();
-      console.log(chalk.yellow('⚠️  No running servers found. Start a server first:'));
-      console.log(chalk.dim('   llamacpp server create <model>'));
-    }
-
-    console.log();
-    console.log(chalk.dim('Quick commands:'));
-    console.log(chalk.dim(`  Status: llamacpp router status`));
-    console.log(chalk.dim(`  Stop:   llamacpp router stop`));
-    console.log(chalk.dim(`  Logs:   tail -f ${config.stderrPath}`));
-  } catch (error) {
-    throw new Error(`Failed to start router: ${(error as Error).message}`);
-  }
-}

package/src/commands/router/status.ts

@@ -1,119 +0,0 @@
-import chalk from 'chalk';
-import { routerManager } from '../../lib/router-manager';
-import { stateManager } from '../../lib/state-manager';
-
-export async function routerStatusCommand(): Promise<void> {
-  try {
-    // Get router status
-    const result = await routerManager.getStatus();
-    if (!result) {
-      console.log(chalk.yellow('Router not configured'));
-      console.log();
-      console.log(chalk.dim('Create and start router:'));
-      console.log(chalk.dim('  llamacpp router start'));
-      return;
-    }
-
-    const { config, status } = result;
-
-    // Calculate uptime if running
-    let uptime = 'N/A';
-    if (status.isRunning && config.lastStarted) {
-      const startTime = new Date(config.lastStarted).getTime();
-      const now = Date.now();
-      const uptimeSeconds = Math.floor((now - startTime) / 1000);
-      const hours = Math.floor(uptimeSeconds / 3600);
-      const minutes = Math.floor((uptimeSeconds % 3600) / 60);
-      const seconds = uptimeSeconds % 60;
-
-      if (hours > 0) {
-        uptime = `${hours}h ${minutes}m`;
-      } else if (minutes > 0) {
-        uptime = `${minutes}m ${seconds}s`;
-      } else {
-        uptime = `${seconds}s`;
-      }
-    }
-
-    // Get running servers
-    const servers = await stateManager.getAllServers();
-    const runningServers = servers.filter(s => s.status === 'running');
-
-    // Display status
-    console.log();
-    console.log(chalk.bold('Router Status'));
-    console.log(chalk.dim('─'.repeat(50)));
-    console.log();
-
-    // Status badge
-    const statusColor = status.isRunning ? chalk.green : chalk.gray;
-    const statusBadge = status.isRunning ? '● RUN' : '○ OFF';
-    console.log(`Status:     ${statusColor(statusBadge)}`);
-
-    if (status.isRunning) {
-      console.log(`PID:        ${status.pid || 'N/A'}`);
-      console.log(`Uptime:     ${uptime}`);
-    }
-
-    console.log(`Port:       ${config.port}`);
-    console.log(`Host:       ${config.host}`);
-    console.log(`Endpoint:   http://${config.host}:${config.port}`);
-    console.log();
-
-    // Available models
-    console.log(chalk.bold('Available Models'));
-    console.log(chalk.dim('─'.repeat(50)));
-    console.log();
-
-    if (runningServers.length === 0) {
-      console.log(chalk.dim('No running servers found'));
-      console.log();
-      console.log(chalk.yellow('⚠️  Start a server first:'));
-      console.log(chalk.dim('   llamacpp server create <model>'));
-    } else {
-      runningServers.forEach(server => {
-        console.log(`  ${chalk.green('●')} ${server.modelName}`);
-        console.log(chalk.dim(`    Port: ${server.port}`));
-        console.log(chalk.dim(`    Backend: http://${server.host}:${server.port}`));
-        console.log();
-      });
-    }
-
-    // Configuration
-    console.log(chalk.bold('Configuration'));
-    console.log(chalk.dim('─'.repeat(50)));
-    console.log();
-    console.log(`Health Check Interval: ${config.healthCheckInterval}ms`);
-    console.log(`Request Timeout:       ${config.requestTimeout}ms`);
-    console.log();
-
-    // System paths
-    console.log(chalk.bold('System Paths'));
-    console.log(chalk.dim('─'.repeat(50)));
-    console.log();
-    console.log(chalk.dim(`Config: ${config.plistPath.replace(config.label + '.plist', 'router.json').replace('LaunchAgents', '.llamacpp')}`));
-    console.log(chalk.dim(`Plist:  ${config.plistPath}`));
-    console.log(chalk.dim(`Stdout: ${config.stdoutPath}`));
-    console.log(chalk.dim(`Stderr: ${config.stderrPath}`));
-    console.log();
-
-    // Quick commands
-    console.log(chalk.bold('Quick Commands'));
-    console.log(chalk.dim('─'.repeat(50)));
-    console.log();
-
-    if (status.isRunning) {
-      console.log(chalk.dim('  Stop:    llamacpp router stop'));
-      console.log(chalk.dim('  Restart: llamacpp router restart'));
-      console.log(chalk.dim(`  Logs:    tail -f ${config.stderrPath}`));
-      console.log(chalk.dim('  Config:  llamacpp router config --port <port> --restart'));
-    } else {
-      console.log(chalk.dim('  Start:   llamacpp router start'));
-      console.log(chalk.dim('  Config:  llamacpp router config --port <port>'));
-      console.log(chalk.dim(`  Logs:    cat ${config.stderrPath}`));
-    }
-    console.log();
-  } catch (error) {
-    throw new Error(`Failed to get router status: ${(error as Error).message}`);
-  }
-}

package/src/commands/router/stop.ts

@@ -1,33 +0,0 @@
-import chalk from 'chalk';
-import { routerManager } from '../../lib/router-manager';
-
-export async function routerStopCommand(): Promise<void> {
-  console.log(chalk.blue('⏹️  Stopping router...'));
-
-  try {
-    // Check if router exists
-    const config = await routerManager.loadConfig();
-    if (!config) {
-      throw new Error('Router configuration not found. Use "llamacpp router start" to create it.');
-    }
-
-    // Check if already stopped
-    if (config.status !== 'running') {
-      console.log(chalk.yellow('⚠️  Router is not running'));
-      return;
-    }
-
-    // Stop router
-    await routerManager.stop();
-
-    // Display success
-    console.log();
-    console.log(chalk.green('✅ Router stopped successfully'));
-    console.log();
-    console.log(chalk.dim('Quick commands:'));
-    console.log(chalk.dim('  Start: llamacpp router start'));
-    console.log(chalk.dim('  Status: llamacpp router status'));
-  } catch (error) {
-    throw new Error(`Failed to stop router: ${(error as Error).message}`);
-  }
-}

package/src/commands/run.ts

@@ -1,233 +0,0 @@
-import chalk from 'chalk';
-import * as readline from 'readline';
-import { stateManager } from '../lib/state-manager';
-import { startCommand } from './start';
-import { statusChecker } from '../lib/status-checker';
-import { ServerConfig } from '../types/server-config';
-
-interface ChatMessage {
-  role: 'system' | 'user' | 'assistant';
-  content: string;
-}
-
-interface ChatCompletionChunk {
-  id: string;
-  object: string;
-  created: number;
-  model: string;
-  choices: Array<{
-    index: number;
-    delta: {
-      role?: string;
-      content?: string;
-    };
-    finish_reason: string | null;
-  }>;
-}
-
-interface RunOptions {
-  message?: string;
-}
-
-export async function runCommand(modelIdentifier: string, options: RunOptions = {}): Promise<void> {
-  await stateManager.initialize();
-
-  // 1. Find or start server
-  let server = await stateManager.findServer(modelIdentifier);
-
-  if (!server) {
-    // Try to resolve as a model name and start it
-    console.log(chalk.blue(`🚀 No running server found. Starting ${modelIdentifier}...\n`));
-    try {
-      await startCommand(modelIdentifier);
-      server = await stateManager.findServer(modelIdentifier);
-      if (!server) {
-        throw new Error('Failed to start server');
-      }
-      console.log(); // Add blank line after start output
-    } catch (error) {
-      throw new Error(`Failed to start server: ${(error as Error).message}`);
-    }
-  }
-
-  // 2. Verify server is running
-  const status = await statusChecker.checkServer(server);
-  if (!status.isRunning) {
-    throw new Error(`Server exists but is not running. Start it with: llamacpp server start ${server.id}`);
-  }
-
-  // 3. If message provided, do one-shot mode
-  if (options.message) {
-    const conversationHistory: ChatMessage[] = [
-      {
-        role: 'user',
-        content: options.message,
-      },
-    ];
-
-    try {
-      await streamChatCompletion(server, conversationHistory);
-      console.log(); // Blank line after response
-      process.exit(0);
-    } catch (error) {
-      console.error(chalk.red(`\n❌ Error: ${(error as Error).message}\n`));
-      process.exit(1);
-    }
-    return;
-  }
-
-  // 4. Start REPL
-  console.log(chalk.green(`💬 Connected to ${server.modelName} (port ${server.port})`));
-  console.log(chalk.dim(`Type your message and press Enter. Use /exit to quit, /clear to reset history, /help for commands.\n`));
-
-  const conversationHistory: ChatMessage[] = [];
-  const rl = readline.createInterface({
-    input: process.stdin,
-    output: process.stdout,
-    prompt: chalk.cyan('You: '),
-  });
-
-  // Handle graceful shutdown
-  const cleanup = () => {
-    rl.close();
-    console.log(chalk.dim('\n\nGoodbye!'));
-    process.exit(0);
-  };
-
-  process.on('SIGINT', cleanup);
-  process.on('SIGTERM', cleanup);
-
-  rl.prompt();
-
-  rl.on('line', async (input: string) => {
-    const line = input.trim();
-
-    // Handle special commands
-    if (line === '/exit' || line === '/quit') {
-      cleanup();
-      return;
-    }
-
-    if (line === '/clear') {
-      conversationHistory.length = 0;
-      console.log(chalk.dim('✓ Conversation history cleared\n'));
-      rl.prompt();
-      return;
-    }
-
-    if (line === '/help') {
-      console.log(chalk.bold('\nAvailable commands:'));
-      console.log(chalk.dim('  /exit, /quit  - Exit the chat'));
-      console.log(chalk.dim('  /clear        - Clear conversation history'));
-      console.log(chalk.dim('  /help         - Show this help message\n'));
-      rl.prompt();
-      return;
-    }
-
-    if (!line) {
-      rl.prompt();
-      return;
-    }
-
-    // Add user message to history
-    conversationHistory.push({
-      role: 'user',
-      content: line,
-    });
-
-    // Send to API and stream response
-    try {
-      await streamChatCompletion(server, conversationHistory);
-      console.log(); // Blank line after response
-    } catch (error) {
-      console.error(chalk.red(`\n❌ Error: ${(error as Error).message}\n`));
-    }
-
-    rl.prompt();
-  });
-
-  rl.on('close', () => {
-    cleanup();
-  });
-}
-
-async function streamChatCompletion(
-  server: ServerConfig,
-  messages: ChatMessage[]
-): Promise<void> {
-  const url = `http://localhost:${server.port}/v1/chat/completions`;
-
-  const response = await fetch(url, {
-    method: 'POST',
-    headers: {
-      'Content-Type': 'application/json',
-    },
-    body: JSON.stringify({
-      model: server.modelName,
-      messages: messages,
-      stream: true,
-      temperature: 0.7,
-    }),
-  });
-
-  if (!response.ok) {
-    const errorText = await response.text();
-    throw new Error(`API request failed (${response.status}): ${errorText}`);
-  }
-
-  if (!response.body) {
-    throw new Error('Response body is null');
-  }
-
-  // Display assistant prefix
-  process.stdout.write(chalk.magenta('Assistant: '));
-
-  let fullResponse = '';
-  const reader = response.body.getReader();
-  const decoder = new TextDecoder();
-
-  try {
-    while (true) {
-      const { done, value } = await reader.read();
-      if (done) break;
-
-      const chunk = decoder.decode(value, { stream: true });
-      const lines = chunk.split('\n').filter((line) => line.trim().startsWith('data:'));
-
-      for (const line of lines) {
-        const data = line.replace(/^data:\s*/, '').trim();
-
-        if (data === '[DONE]') {
-          continue;
-        }
-
-        if (!data) {
-          continue;
-        }
-
-        try {
-          const parsed: ChatCompletionChunk = JSON.parse(data);
-          const content = parsed.choices[0]?.delta?.content;
-
-          if (content) {
-            process.stdout.write(content);
-            fullResponse += content;
-          }
-        } catch (parseError) {
-          // Skip malformed JSON chunks
-          continue;
-        }
-      }
-    }
-  } finally {
-    reader.releaseLock();
-  }
-
-  // Add assistant response to history
-  if (fullResponse) {
-    messages.push({
-      role: 'assistant',
-      content: fullResponse,
-    });
-  }
-}

package/src/commands/search.ts

@@ -1,107 +0,0 @@
-import chalk from 'chalk';
-import Table from 'cli-table3';
-import { modelSearch } from '../lib/model-search';
-import { formatBytes } from '../utils/format-utils';
-
-interface SearchOptions {
-  limit?: number;
-  files?: number | boolean;
-}
-
-export async function searchCommand(query: string, options: SearchOptions): Promise<void> {
-  const limit = options.limit || 20;
-
-  console.log(chalk.blue(`🔍 Searching Hugging Face for: "${query}"\n`));
-
-  try {
-    const results = await modelSearch.searchModels(query, limit);
-
-    if (results.length === 0) {
-      console.log(chalk.yellow('No models found.'));
-      console.log(chalk.dim('Try a different search query or browse: https://huggingface.co/models'));
-      return;
-    }
-
-    const table = new Table({
-      head: ['#', 'MODEL ID', 'DOWNLOADS', 'LIKES'],
-      colWidths: [4, 55, 12, 8],
-    });
-
-    for (let i = 0; i < results.length; i++) {
-      const model = results[i];
-      table.push([
-        chalk.dim((i + 1).toString()),
-        model.modelId,
-        model.downloads.toLocaleString(),
-        model.likes.toString(),
-      ]);
-    }
-
-    console.log(table.toString());
-
-    console.log(chalk.dim(`\nShowing ${results.length} results`));
-    console.log(chalk.dim('\nTo see files in a model:'));
-    console.log(chalk.dim('  llamacpp search "<query>" --files <number>'));
-    console.log(chalk.dim('  Example: llamacpp search "llama 3b" --files 1'));
-    console.log(chalk.dim('\nTo download:'));
-    console.log(chalk.dim('  llamacpp pull <model-id>/<file.gguf>'));
-
-    // Handle --files flag
-    if (options.files !== undefined && options.files !== false) {
-      let selectedIndex: number;
-
-      if (typeof options.files === 'number') {
-        // User specified a number: --files 1
-        selectedIndex = options.files - 1; // Convert to 0-based index
-      } else if (results.length === 1) {
-        // No number specified but only one result
-        selectedIndex = 0;
-      } else {
-        // Multiple results but no number specified
-        console.log(chalk.yellow('\n⚠️  Multiple results found. Specify which one:'));
-        console.log(chalk.dim('  llamacpp search "<query>" --files 1'));
-        return;
-      }
-
-      // Validate index
-      if (selectedIndex < 0 || selectedIndex >= results.length) {
-        console.log(chalk.red(`\n❌ Invalid index. Please specify a number between 1 and ${results.length}`));
-        return;
-      }
-
-      await showModelFiles(results[selectedIndex].modelId, selectedIndex + 1);
-    }
-  } catch (error) {
-    throw new Error(`Search failed: ${(error as Error).message}`);
-  }
-}
-
-async function showModelFiles(modelId: string, index?: number): Promise<void> {
-  const indexPrefix = index ? chalk.dim(`[${index}] `) : '';
-  console.log(chalk.blue(`\n📦 GGUF files in ${indexPrefix}${modelId}:\n`));
-
-  try {
-    const files = await modelSearch.getModelFiles(modelId);
-
-    if (files.length === 0) {
-      console.log(chalk.yellow('No GGUF files found in this model.'));
-      return;
-    }
-
-    const table = new Table({
-      head: ['FILENAME'],
-      colWidths: [70],
-    });
-
-    for (const file of files) {
-      table.push([file]);
-    }
-
-    console.log(table.toString());
-
-    console.log(chalk.dim(`\nTo download:`));
-    console.log(chalk.dim(`  llamacpp pull ${modelId}/${files[0]}`));
-  } catch (error) {
-    console.log(chalk.yellow(`\n⚠️  Could not fetch file list: ${(error as Error).message}`));
-  }
-}