diff --git a/lib/browser-buffer.js b/lib/browser-buffer.js
new file mode 100644
index 0000000..9c61b56
--- /dev/null
+++ b/lib/browser-buffer.js
@@ -0,0 +1,206 @@
+"use strict"
+
+// Minimal Buffer shim for iconv-lite browser usage.
+// Implements only the subset of Buffer API used by iconv-lite codecs.
+// Extends Uint8Array so instanceof checks and standard web APIs work.
+
+var textEncoder = new TextEncoder()
+var textDecoder = new TextDecoder()
+
+class BufferShim extends Uint8Array {
+  // Prevent inherited methods (like map, filter) from creating BufferShim instances.
+  static get [Symbol.species] () { return Uint8Array }
+
+  static alloc (size, fill) {
+    var buf = new BufferShim(size)
+    if (fill !== undefined && fill !== 0) {
+      buf.fill(fill)
+    }
+    return buf
+  }
+
+  static from (arg, encoding) {
+    if (typeof arg === "string") {
+      return BufferShim._fromString(arg, encoding)
+    }
+    // Uint8Array, Array, or other array-like
+    var result = new BufferShim(arg.length)
+    result.set(arg)
+    return result
+  }
+
+  static _fromString (str, encoding) {
+    if (!encoding || encoding === "utf8" || encoding === "utf-8") {
+      var encoded = textEncoder.encode(str)
+      var buf = new BufferShim(encoded.length)
+      buf.set(encoded)
+      return buf
+    }
+
+    if (encoding === "ucs2" || encoding === "ucs-2" ||
+        encoding === "utf16le" || encoding === "utf-16le") {
+      var buf = new BufferShim(str.length * 2)
+      for (var i = 0; i < str.length; i++) {
+        var code = str.charCodeAt(i)
+        buf[i * 2] = code & 0xFF
+        buf[i * 2 + 1] = (code >> 8) & 0xFF
+      }
+      return buf
+    }
+
+    if (encoding === "binary" || encoding === "latin1") {
+      var buf = new BufferShim(str.length)
+      for (var i = 0; i < str.length; i++) {
+        buf[i] = str.charCodeAt(i) & 0xFF
+      }
+      return buf
+    }
+
+    if (encoding === "base64") {
+      // Node.js is lenient with base64 - strip invalid chars and add padding
+      var cleaned = str.replace(/[^A-Za-z0-9+/]/g, "")
+      while (cleaned.length % 4 !== 0) cleaned += "="
+      var binaryStr
+      try { binaryStr = atob(cleaned) } catch (e) { binaryStr = "" }
+      var buf = new BufferShim(binaryStr.length)
+      for (var i = 0; i < binaryStr.length; i++) {
+        buf[i] = binaryStr.charCodeAt(i)
+      }
+      return buf
+    }
+
+    if (encoding === "hex") {
+      var len = str.length >> 1
+      var buf = new BufferShim(len)
+      for (var i = 0; i < len; i++) {
+        buf[i] = parseInt(str.substr(i * 2, 2), 16)
+      }
+      return buf
+    }
+
+    if (encoding === "ascii") {
+      var buf = new BufferShim(str.length)
+      for (var i = 0; i < str.length; i++) {
+        buf[i] = str.charCodeAt(i) & 0x7F
+      }
+      return buf
+    }
+
+    // Fallback to utf8
+    return BufferShim._fromString(str, "utf8")
+  }
+
+  static concat (list) {
+    var totalLength = 0
+    for (var i = 0; i < list.length; i++) {
+      totalLength += list[i].length
+    }
+    var result = new BufferShim(totalLength)
+    var offset = 0
+    for (var i = 0; i < list.length; i++) {
+      result.set(list[i], offset)
+      offset += list[i].length
+    }
+    return result
+  }
+
+  static isBuffer (obj) {
+    return obj instanceof BufferShim
+  }
+
+  toString (encoding) {
+    if (!encoding || encoding === "utf8" || encoding === "utf-8") {
+      return textDecoder.decode(this)
+    }
+
+    if (encoding === "ucs2" || encoding === "ucs-2" ||
+        encoding === "utf16le" || encoding === "utf-16le") {
+      var result = ""
+      var CHUNK = 8192
+      var count = this.length >> 1
+      for (var start = 0; start < count; start += CHUNK) {
+        var end = Math.min(start + CHUNK, count)
+        var codes = new Array(end - start)
+        for (var i = start; i < end; i++) {
+          codes[i - start] = this[i * 2] | (this[i * 2 + 1] << 8)
+        }
+        result += String.fromCharCode.apply(null, codes)
+      }
+      return result
+    }
+
+    if (encoding === "base64") {
+      var binaryStr = ""
+      for (var i = 0; i < this.length; i++) {
+        binaryStr += String.fromCharCode(this[i])
+      }
+      return btoa(binaryStr)
+    }
+
+    if (encoding === "binary" || encoding === "latin1") {
+      var result = ""
+      for (var i = 0; i < this.length; i++) {
+        result += String.fromCharCode(this[i])
+      }
+      return result
+    }
+
+    if (encoding === "hex") {
+      var result = ""
+      for (var i = 0; i < this.length; i++) {
+        result += (this[i] < 16 ? "0" : "") + this[i].toString(16)
+      }
+      return result
+    }
+
+    if (encoding === "ascii") {
+      var result = ""
+      for (var i = 0; i < this.length; i++) {
+        result += String.fromCharCode(this[i] & 0x7F)
+      }
+      return result
+    }
+
+    // Fallback to utf8
+    return textDecoder.decode(this)
+  }
+
+  slice (start, end) {
+    var sliced = Uint8Array.prototype.slice.call(this, start, end)
+    Object.setPrototypeOf(sliced, BufferShim.prototype)
+    return sliced
+  }
+
+  write (str, offset) {
+    // Only used in utf7.js for writing ASCII strings at an offset.
+    // Returns number of bytes written.
+    if (offset === undefined) offset = 0
+    var len = Math.min(str.length, this.length - offset)
+    for (var i = 0; i < len; i++) {
+      this[offset + i] = str.charCodeAt(i) & 0xFF
+    }
+    return len
+  }
+
+  readUInt16LE (offset) {
+    return this[offset] | (this[offset + 1] << 8)
+  }
+
+  writeUInt32LE (value, offset) {
+    this[offset] = value & 0xFF
+    this[offset + 1] = (value >>> 8) & 0xFF
+    this[offset + 2] = (value >>> 16) & 0xFF
+    this[offset + 3] = (value >>> 24) & 0xFF
+    return offset + 4
+  }
+
+  writeUInt32BE (value, offset) {
+    this[offset] = (value >>> 24) & 0xFF
+    this[offset + 1] = (value >>> 16) & 0xFF
+    this[offset + 2] = (value >>> 8) & 0xFF
+    this[offset + 3] = value & 0xFF
+    return offset + 4
+  }
+}
+
+module.exports = { Buffer: BufferShim }
diff --git a/lib/browser-string-decoder.js b/lib/browser-string-decoder.js
new file mode 100644
index 0000000..719298c
--- /dev/null
+++ b/lib/browser-string-decoder.js
@@ -0,0 +1,142 @@
+"use strict"
+
+// Minimal StringDecoder shim for iconv-lite browser usage.
+// Replaces Node.js string_decoder module.
+// Supports only the encodings used by iconv-lite's internal.js: utf8, ucs2, binary, base64, hex.
+
+function BrowserStringDecoder (encoding) {
+  this.encoding = normalizeEncoding(encoding)
+
+  if (this.encoding === "utf8") {
+    this._decoder = new TextDecoder("utf-8", { fatal: false })
+  } else if (this.encoding === "ucs2") {
+    this._lastByte = -1
+  } else if (this.encoding === "base64") {
+    this._remainder = null
+  }
+  // binary, hex, and ascii need no state
+}
+
+function normalizeEncoding (enc) {
+  enc = (enc || "utf8").toLowerCase().replace(/[^a-z0-9]/g, "")
+  if (enc === "utf8") return "utf8"
+  if (enc === "ucs2" || enc === "utf16le") return "ucs2"
+  if (enc === "binary" || enc === "latin1") return "binary"
+  if (enc === "base64") return "base64"
+  if (enc === "hex") return "hex"
+  if (enc === "ascii") return "ascii"
+  return "utf8" // fallback
+}
+
+BrowserStringDecoder.prototype.write = function (buf) {
+  if (!buf || buf.length === 0) return ""
+
+  switch (this.encoding) {
+    case "utf8":
+      return this._decoder.decode(buf, { stream: true })
+
+    case "ucs2": {
+      var str = ""
+      var i = 0
+      if (this._lastByte !== -1) {
+        if (buf.length > 0) {
+          str += String.fromCharCode(this._lastByte | (buf[0] << 8))
+          i = 1
+        }
+        this._lastByte = -1
+      }
+      var end = buf.length
+      if ((end - i) % 2 !== 0) {
+        this._lastByte = buf[end - 1]
+        end--
+      }
+      for (; i < end; i += 2) {
+        str += String.fromCharCode(buf[i] | (buf[i + 1] << 8))
+      }
+      return str
+    }
+
+    case "binary": {
+      var str = ""
+      for (var i = 0; i < buf.length; i++) {
+        str += String.fromCharCode(buf[i])
+      }
+      return str
+    }
+
+    case "base64": {
+      var data = buf
+      if (this._remainder) {
+        var combined = new Uint8Array(this._remainder.length + buf.length)
+        combined.set(this._remainder)
+        combined.set(buf, this._remainder.length)
+        data = combined
+        this._remainder = null
+      }
+      var leftover = data.length % 3
+      if (leftover > 0) {
+        this._remainder = data.slice(data.length - leftover)
+        data = data.slice(0, data.length - leftover)
+      }
+      if (data.length === 0) return ""
+      var binaryStr = ""
+      for (var i = 0; i < data.length; i++) {
+        binaryStr += String.fromCharCode(data[i])
+      }
+      return btoa(binaryStr)
+    }
+
+    case "hex": {
+      var str = ""
+      for (var i = 0; i < buf.length; i++) {
+        str += (buf[i] < 16 ? "0" : "") + buf[i].toString(16)
+      }
+      return str
+    }
+
+    case "ascii": {
+      var str = ""
+      for (var i = 0; i < buf.length; i++) {
+        str += String.fromCharCode(buf[i] & 0x7F)
+      }
+      return str
+    }
+
+    default:
+      return ""
+  }
+}
+
+BrowserStringDecoder.prototype.end = function (buf) {
+  var res = ""
+
+  if (buf && buf.length > 0) {
+    if (this.encoding === "utf8") {
+      // Final call without stream:true to flush
+      res = this._decoder.decode(buf)
+    } else {
+      res = this.write(buf)
+    }
+  } else if (this.encoding === "utf8") {
+    // Flush any remaining incomplete bytes
+    res = this._decoder.decode()
+  }
+
+  // Flush buffered state
+  if (this.encoding === "ucs2" && this._lastByte !== -1) {
+    this._lastByte = -1
+  }
+
+  if (this.encoding === "base64" && this._remainder) {
+    var binaryStr = ""
+    for (var i = 0; i < this._remainder.length; i++) {
+      binaryStr += String.fromCharCode(this._remainder[i])
+    }
+    res += btoa(binaryStr)
+    this._remainder = null
+  }
+
+  return res
+}
+
+module.exports = { StringDecoder: BrowserStringDecoder }
diff --git a/package.json b/package.json
index 2a57357..628c8b0 100644
--- a/package.json
+++ b/package.json
@@ -43,7 +43,9 @@
         "typegen": "node generation/gen-typings.js"
     },
     "browser": {
-        "stream": false
+        "stream": false,
+        "safer-buffer": "./lib/browser-buffer.js",
+        "string_decoder": "./lib/browser-string-decoder.js"
     },
     "devDependencies": {
         "@arethetypeswrong/cli": "^0.17.4",