From 613895afbfc4b27a9f56ff60f8815f325b71ee81 Mon Sep 17 00:00:00 2001
From: clawbot <35+clawbot@noreply.example.org>
Date: Mon, 12 Oct 2026 22:51:00 +0000
Subject: [PATCH] Derive the vault key from the mnemonic's words however they
 are spaced (closes #135)

The long-term key was derived from the mnemonic text exactly as typed,
so a double space, a tab or a trailing newline gave a different key,
and a vault made that way did not open with the same words typed
normally. agehd.DeriveEntropy, which init, vault create, vault import,
unlocker add and every unlock go through, now joins the words with
single spaces and puts them in Unicode NFKD, as BIP39 defines the
seed's input, before hashing them. golang.org/x/text, already an
indirect dependency, becomes a direct one for the NFKD step.

Judgement call: internal/bip39's NewSeed keeps hashing its input as
given, as upstream does; the normalising is in agehd.

Model: opus-5-5
---
 TODO.md                 |  7 +++++++
 go.mod                  |  2 +-
 pkg/agehd/agehd.go      |  9 +++++++--
 pkg/agehd/agehd_test.go | 28 ++++++++++++++++++++++++++++
 4 files changed, 43 insertions(+), 3 deletions(-)

diff --git a/TODO.md b/TODO.md
index 5964487..54fcc47 100644
--- a/TODO.md
+++ b/TODO.md
@@ -18,6 +18,13 @@ https://git.eeqj.de/sneak/secret/milestone/12
 
 # Completed Steps
 
+- 2026-10-12: A mnemonic typed with extra spaces, tabs or line breaks gives the
+  same long-term key as its words separated by single spaces
+  (https://git.eeqj.de/sneak/secret/issues/135). The key was derived from the
+  text exactly as typed, so a vault made with a double or trailing space did not
+  open with the same words typed normally. `agehd.DeriveEntropy`, which every
+  use of the mnemonic goes through, now joins the words with single spaces and
+  puts them in Unicode NFKD, as BIP39 defines, before making the seed.
 - 2026-10-12: The keychain and Secure Enclave unlockers are gone from the code
   and the documents (https://git.eeqj.de/sneak/secret/issues/132). They were
   research: neither works without macOS code signing and entitlements, which
diff --git a/go.mod b/go.mod
index fc3d368..cdc8686 100644
--- a/go.mod
+++ b/go.mod
@@ -19,6 +19,7 @@ require (
 	golang.org/x/crypto v0.38.0
 	golang.org/x/sys v0.33.0
 	golang.org/x/term v0.32.0
+	golang.org/x/text v0.25.0
 )
 
 require (
@@ -31,6 +32,5 @@ require (
 	github.com/mattn/go-isatty v0.0.20 // indirect
 	github.com/pmezard/go-difflib v1.0.0 // indirect
 	github.com/spf13/pflag v1.0.6 // indirect
-	golang.org/x/text v0.25.0 // indirect
 	gopkg.in/yaml.v3 v3.0.1 // indirect
 )
diff --git a/pkg/agehd/agehd.go b/pkg/agehd/agehd.go
index 863ea82..bf64473 100644
--- a/pkg/agehd/agehd.go
+++ b/pkg/agehd/agehd.go
@@ -17,6 +17,7 @@ import (
 	"github.com/btcsuite/btcd/btcutil/hdkeychain"
 	"github.com/btcsuite/btcd/chaincfg"
 	"github.com/btcsuite/btcutil/bech32"
+	"golang.org/x/text/unicode/norm"
 	"sneak.berlin/go/secret/internal/bip39"
 	"sneak.berlin/go/secret/pkg/bip85"
 )
@@ -73,8 +74,12 @@ func IdentityFromEntropy(ent []byte) (*age.X25519Identity, error) {
 // DeriveEntropy derives 32 bytes of application-scoped entropy from the
 // supplied BIP-39 mnemonic and index n using BIP85.
 func DeriveEntropy(mnemonic string, n uint32) ([]byte, error) {
-	// Convert mnemonic to seed
-	seed := bip39.NewSeed(mnemonic, "")
+	// BIP-39 makes the seed from the words joined by single spaces, in
+	// Unicode NFKD. NewSeed hashes its input as given, so the mnemonic is
+	// brought to that form first: the same words, however they were
+	// typed, then give the same seed.
+	sentence := norm.NFKD.String(strings.Join(strings.Fields(mnemonic), " "))
+	seed := bip39.NewSeed(sentence, "")
 
 	// Create master key from seed
 	masterKey, err := hdkeychain.NewMaster(seed, &chaincfg.MainNetParams)
diff --git a/pkg/agehd/agehd_test.go b/pkg/agehd/agehd_test.go
index 7573933..e50ca3f 100644
--- a/pkg/agehd/agehd_test.go
+++ b/pkg/agehd/agehd_test.go
@@ -175,6 +175,34 @@ func TestDeterministicDerivation(t *testing.T) {
 	t.Logf("Index 1: %s", id3.String())
 }
 
+// TestDeriveIdentityIgnoresExtraWhitespace checks that the same words give
+// the same identity however they are spaced, as when a mnemonic is typed or
+// pasted with a double space or a trailing newline.
+func TestDeriveIdentityIgnoresExtraWhitespace(t *testing.T) {
+	t.Parallel()
+
+	want, err := DeriveIdentity(mnemonic, 0)
+	if err != nil {
+		t.Fatalf("derive: %v", err)
+	}
+
+	typedMnemonics := []string{
+		strings.Replace(mnemonic, " ", "  ", 1) + " ",
+		"\t" + strings.ReplaceAll(mnemonic, " ", "\n") + "\n",
+	}
+
+	for _, typed := range typedMnemonics {
+		got, err := DeriveIdentity(typed, 0)
+		if err != nil {
+			t.Fatalf("derive from %q: %v", typed, err)
+		}
+
+		if got.String() != want.String() {
+			t.Errorf("%q derives a different identity from %q", typed, mnemonic)
+		}
+	}
+}
+
 func TestDeterministicXPRVDerivation(t *testing.T) {
 	t.Parallel()
 
-- 
2.54.0