claude-brain 0.27.2 → 0.28.0

package/VERSION

@@ -1 +1 @@
-0.27.2
+0.28.0

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "claude-brain",
-  "version": "0.27.2",
+  "version": "0.28.0",
   "description": "Local development assistant bridging Obsidian vaults with Claude Code via MCP",
   "type": "module",
   "main": "src/index.ts",
@@ -73,10 +73,12 @@
     "hono": "4.11.5",
     "lru-cache": "11.2.5",
     "minisearch": "^6.3.0",
+    "onnxruntime-web": "1.24.1",
     "ora": "9.2.0",
     "pino": "^10.1.1",
     "pino-pretty": "^13.1.3",
     "prompts": "2.4.2",
+    "tiktoken": "1.0.22",
     "tree-sitter-wasms": "0.1.13",
     "web-tree-sitter": "0.24.7",
     "zod": "^4.3.5"

package/src/cli/bin.ts

@@ -44,6 +44,8 @@ function printHelp() {
     ['status', 'Show system overview'],
     ['reindex', 'Rebuild code intelligence index'],
     ['export', 'Export observations (--project, --format md|json)'],
+    ['export-training', 'Export SLM training data (intent|entity|query|knowledge|compress|pattern)'],
+    ['models', 'Manage SLM models (list/download/enable/disable/benchmark/stats)'],
     ['health', 'Run health checks'],
     ['diagnose', 'Run diagnostics'],
     ['version', 'Show version'],
@@ -212,6 +214,18 @@ async function main() {
       break
     }
 
+    case 'export-training': {
+      const { runExportTraining } = await import('./commands/export-training')
+      await runExportTraining()
+      break
+    }
+
+    case 'models': {
+      const { runModels } = await import('./commands/models')
+      await runModels()
+      break
+    }
+
     case 'health': {
       const { runHealthCheck } = await import('@/health')
       await runHealthCheck()

package/src/cli/commands/export-training.ts

@@ -0,0 +1,70 @@
+/**
+ * Export Training Data Command — Phase 1A (SLM Upgrade)
+ * Exports logged classification data as JSONL for model training.
+ *
+ * Usage:
+ *   claude-brain export-training intent
+ *   claude-brain export-training entity --verified-only
+ *   claude-brain export-training compress --limit 500
+ *   claude-brain export-training --stats
+ */
+
+import { parseArgs } from 'citty'
+import { exportTrainingData, getTrainingStats, type TrainingTask } from '@/training/data-store'
+
+const VALID_TASKS: TrainingTask[] = ['intent', 'entity', 'query', 'knowledge', 'compress', 'pattern']
+
+export async function runExportTraining() {
+  const args = parseArgs(process.argv.slice(3), {
+    task: { type: 'positional', description: 'Task to export: intent, entity, query, knowledge, compress, pattern' },
+    verifiedOnly: { type: 'boolean', alias: ['verified-only'], description: 'Only export human-verified examples' },
+    limit: { type: 'string', description: 'Max examples to export' },
+    stats: { type: 'boolean', description: 'Show training data statistics' },
+    format: { type: 'string', default: 'jsonl', description: 'Output format: jsonl or csv' },
+  })
+
+  // Stats mode
+  if (args.stats) {
+    const stats = getTrainingStats()
+    console.log('Training Data Statistics:')
+    console.log('========================')
+    let grandTotal = 0
+    let grandVerified = 0
+    for (const [task, counts] of Object.entries(stats)) {
+      console.log(`  ${task.padEnd(12)} ${String(counts.total).padStart(6)} total  ${String(counts.verified).padStart(6)} verified`)
+      grandTotal += counts.total
+      grandVerified += counts.verified
+    }
+    console.log('  ' + '-'.repeat(40))
+    console.log(`  ${'TOTAL'.padEnd(12)} ${String(grandTotal).padStart(6)} total  ${String(grandVerified).padStart(6)} verified`)
+    return
+  }
+
+  const task = args.task as TrainingTask
+  if (!task || !VALID_TASKS.includes(task)) {
+    console.error(`Error: task must be one of: ${VALID_TASKS.join(', ')}`)
+    console.error('Usage: claude-brain export-training <task> [--verified-only] [--limit N]')
+    console.error('       claude-brain export-training --stats')
+    process.exit(1)
+  }
+
+  const limit = args.limit ? parseInt(args.limit, 10) : undefined
+  const lines = exportTrainingData(task, {
+    verifiedOnly: args.verifiedOnly || false,
+    limit,
+  })
+
+  if (lines.length === 0) {
+    console.error(`No training data found for task "${task}".`)
+    console.error('Training data is collected automatically as Claude Brain processes requests.')
+    process.exit(0)
+  }
+
+  // Output to stdout (pipe-friendly)
+  for (const line of lines) {
+    console.log(line)
+  }
+
+  // Summary to stderr so it doesn't pollute piped output
+  console.error(`\nExported ${lines.length} examples for task "${task}"`)
+}