fix[notask]: harden benchmark servers against RCE, path traversal, injection, DoS, and CSRF

packages/qvac-lib-infer-parakeet/benchmarks/server/src/server.js

@@ -85,9 +85,13 @@ const handleRequest = async (req, res) => {
   }
 
   res.setHeader('Content-Type', 'application/json')
-  res.setHeader('Access-Control-Allow-Origin', '*')
+
+  const origin = req.headers.origin || ''
+  const allowedOrigin = /^https?:\/\/(localhost|127\.0\.0\.1)(:\d+)?$/.test(origin) ? origin : ''
+  res.setHeader('Access-Control-Allow-Origin', allowedOrigin)
   res.setHeader('Access-Control-Allow-Methods', 'GET, POST, OPTIONS')
   res.setHeader('Access-Control-Allow-Headers', 'Content-Type')
+  res.setHeader('Vary', 'Origin')
 
   if (method === HTTP_METHODS.OPTIONS) {
     res.statusCode = 204

packages/qvac-lib-infer-parakeet/benchmarks/server/src/services/runAddon.js

@@ -1,15 +1,36 @@
 'use strict'
 
 const { InferenceArgsSchema } = require('../validation')
-const { spawn } = require('bare-subprocess')
 const logger = require('../utils/logger')
 const fs = require('bare-fs')
 const { Readable } = require('bare-stream')
 const process = require('bare-process')
 const path = require('bare-path')
 
+const ALLOWED_LIBS = [
+  '@qvac/transcription-parakeet'
+]
+
 const loadedModels = new Map()
 
+const ALLOWED_AUDIO_DIRS = [
+  path.resolve('.'),
+  path.resolve('./models'),
+  path.resolve('./examples')
+]
+
+const validateFilePath = (filePath) => {
+  const resolved = path.resolve(filePath)
+  if (!fs.existsSync(resolved)) {
+    throw new Error('File not found')
+  }
+  const isAllowed = ALLOWED_AUDIO_DIRS.some(dir => resolved.startsWith(dir + path.sep) || resolved === dir)
+  if (!isAllowed) {
+    throw new Error('File path is outside allowed directories')
+  }
+  return resolved
+}
+
 const getPackageVersion = (lib) => {
   try {
     const packagePath = require.resolve(`${lib}/package`)
@@ -21,43 +42,6 @@ const getPackageVersion = (lib) => {
   }
 }
 
-const ensurePackage = async (lib, requestedVersion) => {
-  const installed = getPackageVersion(lib)
-
-  // If package is already installed, use it (skip version check for local installs)
-  if (installed) {
-    if (!requestedVersion || installed === requestedVersion) {
-      logger.info(`Using installed ${lib}@${installed}`)
-      return installed
-    }
-    // Version mismatch but package is installed - use installed version with warning
-    logger.warn(`Requested ${lib}@${requestedVersion} but ${installed} is installed. Using installed version.`)
-    return installed
-  }
-
-  // Package not installed - try to install from npm
-  const versionSpec = requestedVersion ? `@${requestedVersion}` : ''
-  logger.info(`Installing ${lib}${versionSpec}...`)
-  await new Promise((resolve, reject) => {
-    const npm = spawn('npm', ['install', `${lib}${versionSpec}`], { stdio: 'inherit' })
-    npm
-      .on('exit', code => code === 0 ? resolve() : reject(new Error(`npm install ${lib}${versionSpec} failed (${code})`)))
-      .on('error', reject)
-  })
-
-  // Try to get version after install, but don't fail if we can't verify
-  // (Bare runtime has different module resolution than Node.js)
-  const newVersion = getPackageVersion(lib)
-  if (newVersion) {
-    logger.info(`Installed ${lib}@${newVersion}`)
-    return newVersion
-  }
-
-  // Package installed but version couldn't be verified - return 'unknown'
-  logger.warn(`Installed ${lib} but couldn't verify version (Bare runtime). Proceeding anyway.`)
-  return 'unknown'
-}
-
 class FakeLoader {
   async start () {}
   async stop () {}
@@ -118,9 +102,13 @@ const runAddon = async (payload) => {
     const { inputs, parakeet, config } =
       InferenceArgsSchema.parse(payload)
 
-    const { lib: parakeetLib, version: parakeetVerReq } = parakeet
+    const { lib: parakeetLib } = parakeet
+
+    if (!ALLOWED_LIBS.includes(parakeetLib)) {
+      throw new Error('Unsupported library: ' + parakeetLib + '. Allowed: ' + ALLOWED_LIBS.join(', '))
+    }
 
-    const parakeetVersion = await ensurePackage(parakeetLib, parakeetVerReq)
+    const parakeetVersion = getPackageVersion(parakeetLib) || 'unknown'
     logger.info(`Loading addon: ${parakeetLib}`)
     const TranscriptionParakeet = require(parakeetLib)
     logger.info('Addon loaded successfully')
@@ -144,6 +132,7 @@ const runAddon = async (payload) => {
       if (!config.path) {
         throw new Error('Model path is required in config')
       }
+      validateFilePath(config.path)
 
       const constructorArgs = {
         loader: new FakeLoader(),
@@ -191,7 +180,8 @@ const runAddon = async (payload) => {
     const runStart = process.hrtime()
 
     for (const audioFilePath of inputs) {
-      const audioBuffer = fs.readFileSync(audioFilePath)
+      const resolvedAudioPath = validateFilePath(audioFilePath)
+      const audioBuffer = fs.readFileSync(resolvedAudioPath)
       const segments = []
 
       let audioStream

packages/qvac-lib-infer-parakeet/benchmarks/server/src/utils/helper.js

@@ -2,20 +2,30 @@
 
 const { ERRORS } = require('./constants')
 const ApiError = require('./ApiError')
+const { Buffer } = require('bare-buffer')
 
 /**
  * Process JSON request body
  * @param {http.IncomingMessage} req
  * @returns {Promise<Object>}
  */
+const MAX_BODY_SIZE = 1 * 1024 * 1024 // 1 MB
+
 const processJsonRequest = async (req) => {
   return new Promise((resolve, reject) => {
-    let body = ''
+    const chunks = []
+    let received = 0
     req.on('data', chunk => {
-      body += chunk.toString()
+      received += chunk.length
+      if (received > MAX_BODY_SIZE) {
+        req.destroy(new ApiError(413, 'Payload too large'))
+        return
+      }
+      chunks.push(chunk)
     })
     req.on('end', () => {
       try {
+        const body = Buffer.concat(chunks, received).toString('utf8')
         if (!body) {
           resolve({})
           return

packages/qvac-lib-infer-whispercpp/benchmarks/server/src/server.js

@@ -87,9 +87,13 @@ const handleRequest = async (req, res) => {
   }
 
   res.setHeader('Content-Type', 'application/json')
-  res.setHeader('Access-Control-Allow-Origin', '*')
+
+  const origin = req.headers.origin || ''
+  const allowedOrigin = /^https?:\/\/(localhost|127\.0\.0\.1)(:\d+)?$/.test(origin) ? origin : ''
+  res.setHeader('Access-Control-Allow-Origin', allowedOrigin)
   res.setHeader('Access-Control-Allow-Methods', 'GET, POST, OPTIONS')
   res.setHeader('Access-Control-Allow-Headers', 'Content-Type')
+  res.setHeader('Vary', 'Origin')
 
   if (method === HTTP_METHODS.OPTIONS) {
     res.statusCode = 204

packages/qvac-lib-infer-whispercpp/benchmarks/server/src/services/runAddon.cli.js

@@ -6,8 +6,28 @@ const logger = require('../utils/logger')
 const fs = require('bare-fs')
 const process = require('bare-process')
 
+const path = require('bare-path')
+
 const WHISPER_CPP_PATH = '/path/to/whisper.cpp/build/bin/whisper-cli'
 
+const ALLOWED_DIRS = [
+  path.resolve('.'),
+  path.resolve('./models'),
+  path.resolve('./examples')
+]
+
+const validatePath = (filePath) => {
+  const resolved = path.resolve(filePath)
+  if (!fs.existsSync(resolved)) {
+    throw new Error('File not found')
+  }
+  const isAllowed = ALLOWED_DIRS.some(dir => resolved.startsWith(dir + path.sep) || resolved === dir)
+  if (!isAllowed) {
+    throw new Error('File path is outside allowed directories')
+  }
+  return resolved
+}
+
 const convertRawToWav = async (rawFilePath, wavFilePath) => {
   return new Promise((resolve, reject) => {
     const args = [
@@ -31,15 +51,17 @@ const convertRawToWav = async (rawFilePath, wavFilePath) => {
 }
 
 const runWhisperCppCli = async (audioFilePath, modelPath) => {
-  const wavFilePath = audioFilePath.replace('.raw', '.wav')
-  await convertRawToWav(audioFilePath, wavFilePath)
+  const resolvedAudio = validatePath(audioFilePath)
+  const resolvedModel = validatePath(modelPath)
+  const wavFilePath = resolvedAudio.replace(/\.raw$/, '.wav')
+  await convertRawToWav(resolvedAudio, wavFilePath)
 
   return new Promise((resolve, reject) => {
     let stdout = ''
     let stderr = ''
 
     const args = [
-      '-m', modelPath,
+      '-m', resolvedModel,
       '-f', wavFilePath,
       '--output-txt',
       '--no-timestamps'

packages/qvac-lib-infer-whispercpp/benchmarks/server/src/services/runAddon.js

@@ -1,15 +1,36 @@
 'use strict'
 
 const { InferenceArgsSchema } = require('../validation')
-const { spawn } = require('bare-subprocess')
 const logger = require('../utils/logger')
 const fs = require('bare-fs')
 const { Readable } = require('bare-stream')
 const process = require('bare-process')
 const path = require('bare-path')
 
+const ALLOWED_LIBS = [
+  '@qvac/transcription-whispercpp'
+]
+
 const loadedModels = new Map()
 
+const ALLOWED_AUDIO_DIRS = [
+  path.resolve('.'),
+  path.resolve('./models'),
+  path.resolve('./examples')
+]
+
+const validateFilePath = (filePath) => {
+  const resolved = path.resolve(filePath)
+  if (!fs.existsSync(resolved)) {
+    throw new Error('File not found')
+  }
+  const isAllowed = ALLOWED_AUDIO_DIRS.some(dir => resolved.startsWith(dir + path.sep) || resolved === dir)
+  if (!isAllowed) {
+    throw new Error('File path is outside allowed directories')
+  }
+  return resolved
+}
+
 const getPackageVersion = (lib) => {
   try {
     const packagePath = require.resolve(`${lib}/package`)
@@ -21,26 +42,6 @@ const getPackageVersion = (lib) => {
   }
 }
 
-const ensurePackage = async (lib, requestedVersion) => {
-  const installed = getPackageVersion(lib)
-  if (installed && (!requestedVersion || installed === requestedVersion)) {
-    return installed
-  }
-  const versionSpec = requestedVersion ? `@${requestedVersion}` : ''
-  logger.info(`Installing ${lib}${versionSpec}...`)
-  await new Promise((resolve, reject) => {
-    const npm = spawn('npm', ['install', `${lib}${versionSpec}`], { stdio: 'inherit' })
-    npm
-      .on('exit', code => code === 0 ? resolve() : reject(new Error(`npm install ${lib}${versionSpec} failed (${code})`)))
-      .on('error', reject)
-  })
-  const newVersion = getPackageVersion(lib)
-  if (!newVersion) {
-    throw new Error(`Failed to verify installation of ${lib}${versionSpec}`)
-  }
-  return newVersion
-}
-
 class FakeLoader {
   async start () {}
   async stop () {}
@@ -70,9 +71,13 @@ const runAddon = async (payload) => {
   const { inputs, whisper, config } =
     InferenceArgsSchema.parse(payload)
 
-  const { lib: whisperLib, version: whisperVerReq } = whisper
+  const { lib: whisperLib } = whisper
+
+  if (!ALLOWED_LIBS.includes(whisperLib)) {
+    throw new Error('Unsupported library: ' + whisperLib + '. Allowed: ' + ALLOWED_LIBS.join(', '))
+  }
 
-  const whisperVersion = await ensurePackage(whisperLib, whisperVerReq)
+  const whisperVersion = getPackageVersion(whisperLib) || 'unknown'
   const TranscriptionWhispercpp = require(whisperLib)
 
   logger.info(`Running addon with ${inputs.length} inputs`)
@@ -93,6 +98,7 @@ const runAddon = async (payload) => {
     if (!config.path) {
       throw new Error('Model path is required in config')
     }
+    validateFilePath(config.path)
 
     const constructorArgs = {
       loader: new FakeLoader(),
@@ -150,7 +156,8 @@ const runAddon = async (payload) => {
   const runStart = process.hrtime()
 
   for (const audioFilePath of inputs) {
-    const audioBuffer = fs.readFileSync(audioFilePath)
+    const resolvedAudioPath = validateFilePath(audioFilePath)
+    const audioBuffer = fs.readFileSync(resolvedAudioPath)
     const segments = []
 
     let audioStream