diff --git a/go.mod b/go.mod
index 64780f1570..f11d88a2f2 100644
--- a/go.mod
+++ b/go.mod
@@ -175,7 +175,7 @@ require (
 	github.com/sergi/go-diff v1.3.2-0.20230802210424-5b0b94c5c0d3
 	github.com/tklauser/go-sysconf v0.3.12 // indirect
 	github.com/tklauser/numcpus v0.6.1 // indirect
-	github.com/ulikunitz/xz v0.5.11 // indirect
+	github.com/ulikunitz/xz v0.5.14 // indirect
 	github.com/valyala/bytebufferpool v1.0.0 // indirect
 	github.com/valyala/fasttemplate v1.2.2 // indirect
 	github.com/xanzy/ssh-agent v0.3.3 // indirect
diff --git a/go.sum b/go.sum
index 214e1a1a6c..235c0badab 100644
--- a/go.sum
+++ b/go.sum
@@ -679,8 +679,8 @@ github.com/tklauser/numcpus v0.6.1/go.mod h1:1XfjsgE2zo8GVw7POkMbHENHzVg3GzmoZ9f
 github.com/tmc/grpc-websocket-proxy v0.0.0-20190109142713-0ad062ec5ee5/go.mod h1:ncp9v5uamzpCO7NfCPTXjqaC+bZgJeR0sMTm6dMHP7U=
 github.com/ulikunitz/xz v0.5.8/go.mod h1:nbz6k7qbPmH4IRqmfOplQw/tblSgqTqBwxkY0oWt/14=
 github.com/ulikunitz/xz v0.5.9/go.mod h1:nbz6k7qbPmH4IRqmfOplQw/tblSgqTqBwxkY0oWt/14=
-github.com/ulikunitz/xz v0.5.11 h1:kpFauv27b6ynzBNT/Xy+1k+fK4WswhN/6PN5WhFAGw8=
-github.com/ulikunitz/xz v0.5.11/go.mod h1:nbz6k7qbPmH4IRqmfOplQw/tblSgqTqBwxkY0oWt/14=
+github.com/ulikunitz/xz v0.5.14 h1:uv/0Bq533iFdnMHZdRBTOlaNMdb1+ZxXIlHDZHIHcvg=
+github.com/ulikunitz/xz v0.5.14/go.mod h1:nbz6k7qbPmH4IRqmfOplQw/tblSgqTqBwxkY0oWt/14=
 github.com/valyala/bytebufferpool v1.0.0 h1:GqA5TC/0021Y/b9FG4Oi9Mr3q7XYx6KllzawFIhcdPw=
 github.com/valyala/bytebufferpool v1.0.0/go.mod h1:6bBcMArwyJ5K/AmCkWv1jt77kVWyCJ6HpOuEn7z0Csc=
 github.com/valyala/fasttemplate v1.2.2 h1:lxLXG0uE3Qnshl9QyaK6XJxMXlQZELvChBOCmQD0Loo=
diff --git a/vendor/github.com/ulikunitz/xz/README.md b/vendor/github.com/ulikunitz/xz/README.md
index 5547185213..56d49275a7 100644
--- a/vendor/github.com/ulikunitz/xz/README.md
+++ b/vendor/github.com/ulikunitz/xz/README.md
@@ -75,3 +75,14 @@ To decompress it use the following command.
 
     $ gxz -d bigfile.xz
 
+## Security & Vulnerabilities
+
+The security policy is documented in [SECURITY.md](SECURITY.md). 
+
+The software is not affected by the supply chain attack on the original xz
+implementation, [CVE-2024-3094](https://nvd.nist.gov/vuln/detail/CVE-2024-3094).
+This implementation doesn't share any files with the original xz implementation
+and no patches or pull requests are accepted without a review.
+
+All security advisories for this project are published under
+[github.com/ulikunitz/xz/security/advisories](https://github.com/ulikunitz/xz/security/advisories?state=published).
diff --git a/vendor/github.com/ulikunitz/xz/SECURITY.md b/vendor/github.com/ulikunitz/xz/SECURITY.md
index 5f7ec01b3b..1bdc88878d 100644
--- a/vendor/github.com/ulikunitz/xz/SECURITY.md
+++ b/vendor/github.com/ulikunitz/xz/SECURITY.md
@@ -6,5 +6,14 @@ Currently the last minor version v0.5.x is supported.
 
 ## Reporting a Vulnerability
 
-Report a vulnerability by creating a Github issue at
-<https://github.com/ulikunitz/xz/issues>. Expect a response in a week.
+You can privately report a vulnerability following this
+[procedure](https://docs.github.com/en/code-security/security-advisories/guidance-on-reporting-and-writing-information-about-vulnerabilities/privately-reporting-a-security-vulnerability#privately-reporting-a-security-vulnerability).
+Alternatively you can create a Github issue at
+<https://github.com/ulikunitz/xz/issues>.
+
+In both cases expect a response in at least 7 days.
+
+## Security Advisories
+
+All security advisories for this project are published under
+[github.com/ulikunitz/xz/security/advisories](https://github.com/ulikunitz/xz/security/advisories?state=published).
diff --git a/vendor/github.com/ulikunitz/xz/TODO.md b/vendor/github.com/ulikunitz/xz/TODO.md
index a3d6f19250..8f9650c13d 100644
--- a/vendor/github.com/ulikunitz/xz/TODO.md
+++ b/vendor/github.com/ulikunitz/xz/TODO.md
@@ -1,9 +1,5 @@
 # TODO list
 
-## Release v0.5.x
-
-1. Support check flag in gxz command.
-
 ## Release v0.6
 
 1. Review encoder and check for lzma improvements under xz.
@@ -86,6 +82,24 @@
 
 ## Log
 
+## 2025-08-28
+
+Release v0.5.14 addresses the security vulnerability CVE-2025-58058. If you put
+bytes in from of a LZMA stream, the header might not be read correctly and
+memory for the dictionary buffer allocated. I have implemented mitigations for
+the problem.
+
+### 2025-08-20
+
+Release v0.5.13 addressed issue #61 regarding handling of multiple WriteClosers
+together. So I added a new package xio with a WriteCloserStack to address the
+issue.
+
+### 2024-04-03
+
+Release v0.5.12 updates README.md and SECURITY.md to address the supply chain
+attack on the original xz implementation.
+
 ### 2022-12-12
 
 Matt Dantay (@bodgit) reported an issue with the LZMA reader. The implementation
@@ -99,7 +113,7 @@ it.
 
 Mituo Heijo has fuzzed xz and found a bug in the function readIndexBody. The
 function allocated a slice of records immediately after reading the value
-without further checks. Sincex the number has been too large the make function
+without further checks. Since the number has been too large the make function
 did panic. The fix is to check the number against the expected number of records
 before allocating the records.
 
diff --git a/vendor/github.com/ulikunitz/xz/lzma/header.go b/vendor/github.com/ulikunitz/xz/lzma/header.go
index 1ae7d80cab..34aa097e15 100644
--- a/vendor/github.com/ulikunitz/xz/lzma/header.go
+++ b/vendor/github.com/ulikunitz/xz/lzma/header.go
@@ -60,36 +60,36 @@ const noHeaderSize uint64 = 1<<64 - 1
 // HeaderLen provides the length of the LZMA file header.
 const HeaderLen = 13
 
-// header represents the header of an LZMA file.
-type header struct {
-	properties Properties
-	dictCap    int
-	// uncompressed size; negative value if no size is given
-	size int64
+// Header represents the Header of an LZMA file.
+type Header struct {
+	Properties Properties
+	DictSize   uint32
+	// uncompressed Size; negative value if no Size is given
+	Size int64
 }
 
 // marshalBinary marshals the header.
-func (h *header) marshalBinary() (data []byte, err error) {
-	if err = h.properties.verify(); err != nil {
+func (h *Header) marshalBinary() (data []byte, err error) {
+	if err = h.Properties.verify(); err != nil {
 		return nil, err
 	}
-	if !(0 <= h.dictCap && int64(h.dictCap) <= MaxDictCap) {
+	if !(h.DictSize <= MaxDictCap) {
 		return nil, fmt.Errorf("lzma: DictCap %d out of range",
-			h.dictCap)
+			h.DictSize)
 	}
 
 	data = make([]byte, 13)
 
 	// property byte
-	data[0] = h.properties.Code()
+	data[0] = h.Properties.Code()
 
 	// dictionary capacity
-	putUint32LE(data[1:5], uint32(h.dictCap))
+	putUint32LE(data[1:5], uint32(h.DictSize))
 
 	// uncompressed size
 	var s uint64
-	if h.size > 0 {
-		s = uint64(h.size)
+	if h.Size > 0 {
+		s = uint64(h.Size)
 	} else {
 		s = noHeaderSize
 	}
@@ -99,20 +99,20 @@ func (h *header) marshalBinary() (data []byte, err error) {
 }
 
 // unmarshalBinary unmarshals the header.
-func (h *header) unmarshalBinary(data []byte) error {
+func (h *Header) unmarshalBinary(data []byte) error {
 	if len(data) != HeaderLen {
 		return errors.New("lzma.unmarshalBinary: data has wrong length")
 	}
 
 	// properties
 	var err error
-	if h.properties, err = PropertiesForCode(data[0]); err != nil {
+	if h.Properties, err = PropertiesForCode(data[0]); err != nil {
 		return err
 	}
 
 	// dictionary capacity
-	h.dictCap = int(uint32LE(data[1:]))
-	if h.dictCap < 0 {
+	h.DictSize = uint32LE(data[1:])
+	if int(h.DictSize) < 0 {
 		return errors.New(
 			"LZMA header: dictionary capacity exceeds maximum " +
 				"integer")
@@ -121,10 +121,10 @@ func (h *header) unmarshalBinary(data []byte) error {
 	// uncompressed size
 	s := uint64LE(data[5:])
 	if s == noHeaderSize {
-		h.size = -1
+		h.Size = -1
 	} else {
-		h.size = int64(s)
-		if h.size < 0 {
+		h.Size = int64(s)
+		if h.Size < 0 {
 			return errors.New(
 				"LZMA header: uncompressed size " +
 					"out of int64 range")
@@ -134,9 +134,9 @@ func (h *header) unmarshalBinary(data []byte) error {
 	return nil
 }
 
-// validDictCap checks whether the dictionary capacity is correct. This
+// validDictSize checks whether the dictionary capacity is correct. This
 // is used to weed out wrong file headers.
-func validDictCap(dictcap int) bool {
+func validDictSize(dictcap int) bool {
 	if int64(dictcap) == MaxDictCap {
 		return true
 	}
@@ -155,13 +155,16 @@ func validDictCap(dictcap int) bool {
 // dictionary sizes of 2^n or 2^n+2^(n-1) with n >= 10 or 2^32-1. If
 // there is an explicit size it must not exceed 256 GiB. The length of
 // the data argument must be HeaderLen.
+//
+// This function should be disregarded because there is no guarantee that LZMA
+// files follow the constraints.
 func ValidHeader(data []byte) bool {
-	var h header
+	var h Header
 	if err := h.unmarshalBinary(data); err != nil {
 		return false
 	}
-	if !validDictCap(h.dictCap) {
+	if !validDictSize(int(h.DictSize)) {
 		return false
 	}
-	return h.size < 0 || h.size <= 1<<38
+	return h.Size < 0 || h.Size <= 1<<38
 }
diff --git a/vendor/github.com/ulikunitz/xz/lzma/reader.go b/vendor/github.com/ulikunitz/xz/lzma/reader.go
index ae911c3893..b656c26187 100644
--- a/vendor/github.com/ulikunitz/xz/lzma/reader.go
+++ b/vendor/github.com/ulikunitz/xz/lzma/reader.go
@@ -6,25 +6,32 @@
 // Reader and Writer support the classic LZMA format. Reader2 and
 // Writer2 support the decoding and encoding of LZMA2 streams.
 //
-// The package is written completely in Go and doesn't rely on any external
+// The package is written completely in Go and does not rely on any external
 // library.
 package lzma
 
 import (
 	"errors"
+	"fmt"
 	"io"
 )
 
 // ReaderConfig stores the parameters for the reader of the classic LZMA
 // format.
 type ReaderConfig struct {
+	// Since v0.5.14 this parameter sets an upper limit for a .lzma file's
+	// dictionary size. This helps to mitigate problems with mangled
+	// headers.
 	DictCap int
 }
 
 // fill converts the zero values of the configuration to the default values.
 func (c *ReaderConfig) fill() {
 	if c.DictCap == 0 {
-		c.DictCap = 8 * 1024 * 1024
+		// set an upper limit of 2 GB for dictionary capacity to address
+		// the zero prefix security issue.
+		c.DictCap = 1 << 31
+		// original: c.DictCap = 8 * 1024 * 1024
 	}
 }
 
@@ -39,10 +46,33 @@ func (c *ReaderConfig) Verify() error {
 }
 
 // Reader provides a reader for LZMA files or streams.
+//
+// # Security concerns
+//
+// Note that LZMA format doesn't support a magic marker in the header. So
+// [NewReader] cannot determine whether it reads the actual header. For instance
+// the LZMA stream might have a zero byte in front of the reader, leading to
+// larger dictionary sizes and file sizes. The code will detect later that there
+// are problems with the stream, but the dictionary has already been allocated
+// and this might consume a lot of memory.
+//
+// Version 0.5.14 introduces built-in mitigations:
+//
+//   - The [ReaderConfig] DictCap field is now interpreted as a limit for the
+//     dictionary size.
+//   - The default is 2 Gigabytes (2^31 bytes).
+//   - Users can check with the [Reader.Header] method what the actual values are in
+//     their LZMA files and set a smaller limit using [ReaderConfig].
+//   - The dictionary size doesn't exceed the larger of the file size and
+//     the minimum dictionary size. This is another measure to prevent huge
+//     memory allocations for the dictionary.
+//   - The code supports stream sizes only up to a pebibyte (1024^5).
 type Reader struct {
-	lzma io.Reader
-	h    header
-	d    *decoder
+	lzma   io.Reader
+	header Header
+	// headerOrig stores the original header read from the stream.
+	headerOrig Header
+	d          *decoder
 }
 
 // NewReader creates a new reader for an LZMA stream using the classic
@@ -51,8 +81,37 @@ func NewReader(lzma io.Reader) (r *Reader, err error) {
 	return ReaderConfig{}.NewReader(lzma)
 }
 
+// ErrDictSize reports about an error of the dictionary size.
+type ErrDictSize struct {
+	ConfigDictCap  int
+	HeaderDictSize uint32
+	Message        string
+}
+
+// Error returns the error message.
+func (e *ErrDictSize) Error() string {
+	return e.Message
+}
+
+func newErrDictSize(messageformat string,
+	configDictCap int, headerDictSize uint32,
+	args ...interface{}) *ErrDictSize {
+	newArgs := make([]interface{}, len(args)+2)
+	newArgs[0] = configDictCap
+	newArgs[1] = headerDictSize
+	copy(newArgs[2:], args)
+	return &ErrDictSize{
+		ConfigDictCap:  configDictCap,
+		HeaderDictSize: headerDictSize,
+		Message:        fmt.Sprintf(messageformat, newArgs...),
+	}
+}
+
+// We support only files not larger than 1 << 50 bytes (a pebibyte, 1024^5).
+const maxStreamSize = 1 << 50
+
 // NewReader creates a new reader for an LZMA stream in the classic
-// format. The function reads and verifies the the header of the LZMA
+// format. The function reads and verifies the header of the LZMA
 // stream.
 func (c ReaderConfig) NewReader(lzma io.Reader) (r *Reader, err error) {
 	if err = c.Verify(); err != nil {
@@ -66,29 +125,63 @@ func (c ReaderConfig) NewReader(lzma io.Reader) (r *Reader, err error) {
 		return nil, err
 	}
 	r = &Reader{lzma: lzma}
-	if err = r.h.unmarshalBinary(data); err != nil {
+	if err = r.header.unmarshalBinary(data); err != nil {
 		return nil, err
 	}
-	if r.h.dictCap < MinDictCap {
-		r.h.dictCap = MinDictCap
+	r.headerOrig = r.header
+	dictSize := int64(r.header.DictSize)
+	if int64(c.DictCap) < dictSize {
+		return nil, newErrDictSize(
+			"lzma: header dictionary size %[2]d exceeds configured dictionary capacity %[1]d",
+			c.DictCap, uint32(dictSize),
+		)
+	}
+	if dictSize < MinDictCap {
+		dictSize = MinDictCap
+	}
+	// original code: disabled this because there is no point in increasing
+	// the dictionary above what is stated in the file.
+	/*
+		if int64(c.DictCap) > int64(dictSize) {
+			dictSize = int64(c.DictCap)
+		}
+	*/
+	size := r.header.Size
+	if size >= 0 && size < dictSize {
+		dictSize = size
 	}
-	dictCap := r.h.dictCap
-	if c.DictCap > dictCap {
-		dictCap = c.DictCap
+	// Protect against modified or malicious headers.
+	if size > maxStreamSize {
+		return nil, fmt.Errorf(
+			"lzma: stream size %d exceeds a pebibyte (1024^5)",
+			size)
 	}
+	if dictSize < MinDictCap {
+		dictSize = MinDictCap
+	}
+
+	r.header.DictSize = uint32(dictSize)
 
-	state := newState(r.h.properties)
-	dict, err := newDecoderDict(dictCap)
+	state := newState(r.header.Properties)
+	dict, err := newDecoderDict(int(dictSize))
 	if err != nil {
 		return nil, err
 	}
-	r.d, err = newDecoder(ByteReader(lzma), state, dict, r.h.size)
+	r.d, err = newDecoder(ByteReader(lzma), state, dict, r.header.Size)
 	if err != nil {
 		return nil, err
 	}
 	return r, nil
 }
 
+// Header returns the header as read from the LZMA stream. It is intended to
+// allow the user to understand what parameters are typically provided in the
+// headers of the LZMA files and set the DictCap field in [ReaderConfig]
+// accordingly.
+func (r *Reader) Header() (h Header, ok bool) {
+	return r.headerOrig, r.d != nil
+}
+
 // EOSMarker indicates that an EOS marker has been encountered.
 func (r *Reader) EOSMarker() bool {
 	return r.d.eosMarker
diff --git a/vendor/github.com/ulikunitz/xz/lzma/writer.go b/vendor/github.com/ulikunitz/xz/lzma/writer.go
index e8f89811d3..f73bb73f28 100644
--- a/vendor/github.com/ulikunitz/xz/lzma/writer.go
+++ b/vendor/github.com/ulikunitz/xz/lzma/writer.go
@@ -96,21 +96,21 @@ func (c *WriterConfig) Verify() error {
 }
 
 // header returns the header structure for this configuration.
-func (c *WriterConfig) header() header {
-	h := header{
-		properties: *c.Properties,
-		dictCap:    c.DictCap,
-		size:       -1,
+func (c *WriterConfig) header() Header {
+	h := Header{
+		Properties: *c.Properties,
+		DictSize:   uint32(c.DictCap),
+		Size:       -1,
 	}
 	if c.SizeInHeader {
-		h.size = c.Size
+		h.Size = c.Size
 	}
 	return h
 }
 
 // Writer writes an LZMA stream in the classic format.
 type Writer struct {
-	h   header
+	h   Header
 	bw  io.ByteWriter
 	buf *bufio.Writer
 	e   *encoder
@@ -130,12 +130,12 @@ func (c WriterConfig) NewWriter(lzma io.Writer) (w *Writer, err error) {
 		w.buf = bufio.NewWriter(lzma)
 		w.bw = w.buf
 	}
-	state := newState(w.h.properties)
-	m, err := c.Matcher.new(w.h.dictCap)
+	state := newState(w.h.Properties)
+	m, err := c.Matcher.new(int(w.h.DictSize))
 	if err != nil {
 		return nil, err
 	}
-	dict, err := newEncoderDict(w.h.dictCap, c.BufSize, m)
+	dict, err := newEncoderDict(int(w.h.DictSize), c.BufSize, m)
 	if err != nil {
 		return nil, err
 	}
@@ -171,8 +171,8 @@ func (w *Writer) writeHeader() error {
 
 // Write puts data into the Writer.
 func (w *Writer) Write(p []byte) (n int, err error) {
-	if w.h.size >= 0 {
-		m := w.h.size
+	if w.h.Size >= 0 {
+		m := w.h.Size
 		m -= w.e.Compressed() + int64(w.e.dict.Buffered())
 		if m < 0 {
 			m = 0
@@ -192,9 +192,9 @@ func (w *Writer) Write(p []byte) (n int, err error) {
 // Close closes the writer stream. It ensures that all data from the
 // buffer will be compressed and the LZMA stream will be finished.
 func (w *Writer) Close() error {
-	if w.h.size >= 0 {
+	if w.h.Size >= 0 {
 		n := w.e.Compressed() + int64(w.e.dict.Buffered())
-		if n != w.h.size {
+		if n != w.h.Size {
 			return errSize
 		}
 	}
diff --git a/vendor/modules.txt b/vendor/modules.txt
index 0e52af712e..cd57a35508 100644
--- a/vendor/modules.txt
+++ b/vendor/modules.txt
@@ -680,7 +680,7 @@ github.com/tklauser/go-sysconf
 # github.com/tklauser/numcpus v0.6.1
 ## explicit; go 1.13
 github.com/tklauser/numcpus
-# github.com/ulikunitz/xz v0.5.11
+# github.com/ulikunitz/xz v0.5.14
 ## explicit; go 1.12
 github.com/ulikunitz/xz
 github.com/ulikunitz/xz/internal/hash