diff --git a/vendor.conf b/vendor.conf
index 7ebddd06..8ef953a1 100644
--- a/vendor.conf
+++ b/vendor.conf
@@ -37,3 +37,4 @@ gopkg.in/check.v1 64131543e7896d5bcc6bd5a76287eb75ea96c673
 gopkg.in/square/go-jose.v1 40d457b439244b546f023d056628e5184136899b
 gopkg.in/yaml.v2 bef53efd0c76e49e6de55ead051f886bea7e9420
 rsc.io/letsencrypt e770c10b0f1a64775ae91d240407ce00d1a5bdeb https://github.com/dmcgowan/letsencrypt.git
+github.com/opencontainers/go-digest a6d0ee40d4207ea02364bd3b9e8e77b9159ba1eb
diff --git a/vendor/github.com/opencontainers/go-digest/algorithm.go b/vendor/github.com/opencontainers/go-digest/algorithm.go
new file mode 100644
index 00000000..a3c44801
--- /dev/null
+++ b/vendor/github.com/opencontainers/go-digest/algorithm.go
@@ -0,0 +1,144 @@
+package digest
+
+import (
+	"crypto"
+	"fmt"
+	"hash"
+	"io"
+)
+
+// Algorithm identifies and implementation of a digester by an identifier.
+// Note the that this defines both the hash algorithm used and the string
+// encoding.
+type Algorithm string
+
+// supported digest types
+const (
+	SHA256 Algorithm = "sha256" // sha256 with hex encoding
+	SHA384 Algorithm = "sha384" // sha384 with hex encoding
+	SHA512 Algorithm = "sha512" // sha512 with hex encoding
+
+	// Canonical is the primary digest algorithm used with the distribution
+	// project. Other digests may be used but this one is the primary storage
+	// digest.
+	Canonical = SHA256
+)
+
+var (
+	// TODO(stevvooe): Follow the pattern of the standard crypto package for
+	// registration of digests. Effectively, we are a registerable set and
+	// common symbol access.
+
+	// algorithms maps values to hash.Hash implementations. Other algorithms
+	// may be available but they cannot be calculated by the digest package.
+	algorithms = map[Algorithm]crypto.Hash{
+		SHA256: crypto.SHA256,
+		SHA384: crypto.SHA384,
+		SHA512: crypto.SHA512,
+	}
+)
+
+// Available returns true if the digest type is available for use. If this
+// returns false, Digester and Hash will return nil.
+func (a Algorithm) Available() bool {
+	h, ok := algorithms[a]
+	if !ok {
+		return false
+	}
+
+	// check availability of the hash, as well
+	return h.Available()
+}
+
+func (a Algorithm) String() string {
+	return string(a)
+}
+
+// Size returns number of bytes returned by the hash.
+func (a Algorithm) Size() int {
+	h, ok := algorithms[a]
+	if !ok {
+		return 0
+	}
+	return h.Size()
+}
+
+// Set implemented to allow use of Algorithm as a command line flag.
+func (a *Algorithm) Set(value string) error {
+	if value == "" {
+		*a = Canonical
+	} else {
+		// just do a type conversion, support is queried with Available.
+		*a = Algorithm(value)
+	}
+
+	if !a.Available() {
+		return ErrDigestUnsupported
+	}
+
+	return nil
+}
+
+// Digester returns a new digester for the specified algorithm. If the algorithm
+// does not have a digester implementation, nil will be returned. This can be
+// checked by calling Available before calling Digester.
+func (a Algorithm) Digester() Digester {
+	return &digester{
+		alg:  a,
+		hash: a.Hash(),
+	}
+}
+
+// Hash returns a new hash as used by the algorithm. If not available, the
+// method will panic. Check Algorithm.Available() before calling.
+func (a Algorithm) Hash() hash.Hash {
+	if !a.Available() {
+		// Empty algorithm string is invalid
+		if a == "" {
+			panic(fmt.Sprintf("empty digest algorithm, validate before calling Algorithm.Hash()"))
+		}
+
+		// NOTE(stevvooe): A missing hash is usually a programming error that
+		// must be resolved at compile time. We don't import in the digest
+		// package to allow users to choose their hash implementation (such as
+		// when using stevvooe/resumable or a hardware accelerated package).
+		//
+		// Applications that may want to resolve the hash at runtime should
+		// call Algorithm.Available before call Algorithm.Hash().
+		panic(fmt.Sprintf("%v not available (make sure it is imported)", a))
+	}
+
+	return algorithms[a].New()
+}
+
+// FromReader returns the digest of the reader using the algorithm.
+func (a Algorithm) FromReader(rd io.Reader) (Digest, error) {
+	digester := a.Digester()
+
+	if _, err := io.Copy(digester.Hash(), rd); err != nil {
+		return "", err
+	}
+
+	return digester.Digest(), nil
+}
+
+// FromBytes digests the input and returns a Digest.
+func (a Algorithm) FromBytes(p []byte) Digest {
+	digester := a.Digester()
+
+	if _, err := digester.Hash().Write(p); err != nil {
+		// Writes to a Hash should never fail. None of the existing
+		// hash implementations in the stdlib or hashes vendored
+		// here can return errors from Write. Having a panic in this
+		// condition instead of having FromBytes return an error value
+		// avoids unnecessary error handling paths in all callers.
+		panic("write to hash function returned error: " + err.Error())
+	}
+
+	return digester.Digest()
+}
+
+// FromString digests the string input and returns a Digest.
+func (a Algorithm) FromString(s string) Digest {
+	return a.FromBytes([]byte(s))
+}
diff --git a/vendor/github.com/opencontainers/go-digest/digest.go b/vendor/github.com/opencontainers/go-digest/digest.go
new file mode 100644
index 00000000..7c66c30c
--- /dev/null
+++ b/vendor/github.com/opencontainers/go-digest/digest.go
@@ -0,0 +1,140 @@
+package digest
+
+import (
+	"fmt"
+	"hash"
+	"io"
+	"regexp"
+	"strings"
+)
+
+// Digest allows simple protection of hex formatted digest strings, prefixed
+// by their algorithm. Strings of type Digest have some guarantee of being in
+// the correct format and it provides quick access to the components of a
+// digest string.
+//
+// The following is an example of the contents of Digest types:
+//
+// 	sha256:7173b809ca12ec5dee4506cd86be934c4596dd234ee82c0662eac04a8c2c71dc
+//
+// This allows to abstract the digest behind this type and work only in those
+// terms.
+type Digest string
+
+// NewDigest returns a Digest from alg and a hash.Hash object.
+func NewDigest(alg Algorithm, h hash.Hash) Digest {
+	return NewDigestFromBytes(alg, h.Sum(nil))
+}
+
+// NewDigestFromBytes returns a new digest from the byte contents of p.
+// Typically, this can come from hash.Hash.Sum(...) or xxx.SumXXX(...)
+// functions. This is also useful for rebuilding digests from binary
+// serializations.
+func NewDigestFromBytes(alg Algorithm, p []byte) Digest {
+	return Digest(fmt.Sprintf("%s:%x", alg, p))
+}
+
+// NewDigestFromHex returns a Digest from alg and a the hex encoded digest.
+func NewDigestFromHex(alg, hex string) Digest {
+	return Digest(fmt.Sprintf("%s:%s", alg, hex))
+}
+
+// DigestRegexp matches valid digest types.
+var DigestRegexp = regexp.MustCompile(`[a-zA-Z0-9-_+.]+:[a-fA-F0-9]+`)
+
+// DigestRegexpAnchored matches valid digest types, anchored to the start and end of the match.
+var DigestRegexpAnchored = regexp.MustCompile(`^` + DigestRegexp.String() + `$`)
+
+var (
+	// ErrDigestInvalidFormat returned when digest format invalid.
+	ErrDigestInvalidFormat = fmt.Errorf("invalid checksum digest format")
+
+	// ErrDigestInvalidLength returned when digest has invalid length.
+	ErrDigestInvalidLength = fmt.Errorf("invalid checksum digest length")
+
+	// ErrDigestUnsupported returned when the digest algorithm is unsupported.
+	ErrDigestUnsupported = fmt.Errorf("unsupported digest algorithm")
+)
+
+// Parse parses s and returns the validated digest object. An error will
+// be returned if the format is invalid.
+func Parse(s string) (Digest, error) {
+	d := Digest(s)
+	return d, d.Validate()
+}
+
+// FromReader consumes the content of rd until io.EOF, returning canonical digest.
+func FromReader(rd io.Reader) (Digest, error) {
+	return Canonical.FromReader(rd)
+}
+
+// FromBytes digests the input and returns a Digest.
+func FromBytes(p []byte) Digest {
+	return Canonical.FromBytes(p)
+}
+
+// FromString digests the input and returns a Digest.
+func FromString(s string) Digest {
+	return Canonical.FromString(s)
+}
+
+// Validate checks that the contents of d is a valid digest, returning an
+// error if not.
+func (d Digest) Validate() error {
+	s := string(d)
+
+	i := strings.Index(s, ":")
+
+	// validate i then run through regexp
+	if i < 0 || i+1 == len(s) || !DigestRegexpAnchored.MatchString(s) {
+		return ErrDigestInvalidFormat
+	}
+
+	algorithm := Algorithm(s[:i])
+	if !algorithm.Available() {
+		return ErrDigestUnsupported
+	}
+
+	// Digests much always be hex-encoded, ensuring that their hex portion will
+	// always be size*2
+	if algorithm.Size()*2 != len(s[i+1:]) {
+		return ErrDigestInvalidLength
+	}
+
+	return nil
+}
+
+// Algorithm returns the algorithm portion of the digest. This will panic if
+// the underlying digest is not in a valid format.
+func (d Digest) Algorithm() Algorithm {
+	return Algorithm(d[:d.sepIndex()])
+}
+
+// Verifier returns a writer object that can be used to verify a stream of
+// content against the digest. If the digest is invalid, the method will panic.
+func (d Digest) Verifier() Verifier {
+	return hashVerifier{
+		hash:   d.Algorithm().Hash(),
+		digest: d,
+	}
+}
+
+// Hex returns the hex digest portion of the digest. This will panic if the
+// underlying digest is not in a valid format.
+func (d Digest) Hex() string {
+	return string(d[d.sepIndex()+1:])
+}
+
+func (d Digest) String() string {
+	return string(d)
+}
+
+func (d Digest) sepIndex() int {
+	i := strings.Index(string(d), ":")
+
+	if i < 0 {
+		panic(fmt.Sprintf("no ':' separator in digest %q", d))
+	}
+
+	return i
+}
diff --git a/vendor/github.com/opencontainers/go-digest/digester.go b/vendor/github.com/opencontainers/go-digest/digester.go
new file mode 100644
index 00000000..918a3f91
--- /dev/null
+++ b/vendor/github.com/opencontainers/go-digest/digester.go
@@ -0,0 +1,25 @@
+package digest
+
+import "hash"
+
+// Digester calculates the digest of written data. Writes should go directly
+// to the return value of Hash, while calling Digest will return the current
+// value of the digest.
+type Digester interface {
+	Hash() hash.Hash // provides direct access to underlying hash instance.
+	Digest() Digest
+}
+
+// digester provides a simple digester definition that embeds a hasher.
+type digester struct {
+	alg  Algorithm
+	hash hash.Hash
+}
+
+func (d *digester) Hash() hash.Hash {
+	return d.hash
+}
+
+func (d *digester) Digest() Digest {
+	return NewDigest(d.alg, d.hash)
+}
diff --git a/vendor/github.com/opencontainers/go-digest/doc.go b/vendor/github.com/opencontainers/go-digest/doc.go
new file mode 100644
index 00000000..f64b0db3
--- /dev/null
+++ b/vendor/github.com/opencontainers/go-digest/doc.go
@@ -0,0 +1,42 @@
+// Package digest provides a generalized type to opaquely represent message
+// digests and their operations within the registry. The Digest type is
+// designed to serve as a flexible identifier in a content-addressable system.
+// More importantly, it provides tools and wrappers to work with
+// hash.Hash-based digests with little effort.
+//
+// Basics
+//
+// The format of a digest is simply a string with two parts, dubbed the
+// "algorithm" and the "digest", separated by a colon:
+//
+// 	<algorithm>:<digest>
+//
+// An example of a sha256 digest representation follows:
+//
+// 	sha256:7173b809ca12ec5dee4506cd86be934c4596dd234ee82c0662eac04a8c2c71dc
+//
+// In this case, the string "sha256" is the algorithm and the hex bytes are
+// the "digest".
+//
+// Because the Digest type is simply a string, once a valid Digest is
+// obtained, comparisons are cheap, quick and simple to express with the
+// standard equality operator.
+//
+// Verification
+//
+// The main benefit of using the Digest type is simple verification against a
+// given digest. The Verifier interface, modeled after the stdlib hash.Hash
+// interface, provides a common write sink for digest verification. After
+// writing is complete, calling the Verifier.Verified method will indicate
+// whether or not the stream of bytes matches the target digest.
+//
+// Missing Features
+//
+// In addition to the above, we intend to add the following features to this
+// package:
+//
+// 1. A Digester type that supports write sink digest calculation.
+//
+// 2. Suspend and resume of ongoing digest calculations to support efficient digest verification in the registry.
+//
+package digest
diff --git a/vendor/github.com/opencontainers/go-digest/verifiers.go b/vendor/github.com/opencontainers/go-digest/verifiers.go
new file mode 100644
index 00000000..f1db6cda
--- /dev/null
+++ b/vendor/github.com/opencontainers/go-digest/verifiers.go
@@ -0,0 +1,31 @@
+package digest
+
+import (
+	"hash"
+	"io"
+)
+
+// Verifier presents a general verification interface to be used with message
+// digests and other byte stream verifications. Users instantiate a Verifier
+// from one of the various methods, write the data under test to it then check
+// the result with the Verified method.
+type Verifier interface {
+	io.Writer
+
+	// Verified will return true if the content written to Verifier matches
+	// the digest.
+	Verified() bool
+}
+
+type hashVerifier struct {
+	digest Digest
+	hash   hash.Hash
+}
+
+func (hv hashVerifier) Write(p []byte) (n int, err error) {
+	return hv.hash.Write(p)
+}
+
+func (hv hashVerifier) Verified() bool {
+	return hv.digest == NewDigest(hv.digest.Algorithm(), hv.hash)
+}