diff --git a/dev/check-spelling.mjs b/dev/check-spelling.mjs
index 733949636..cf07ab1c9 100644
--- a/dev/check-spelling.mjs
+++ b/dev/check-spelling.mjs
@@ -1,7 +1,8 @@
 #!/usr/bin/env node
 
 /**
- * Reports CSpell findings on lines added by a Git diff.
+ * Reports CSpell findings on lines added by a Git diff, and dictionary entries
+ * added out of alphabetical order.
  *
  * Usage: node dev/check-spelling.mjs --base <revision> [--format text|json]
  *
@@ -10,6 +11,7 @@
  */
 
 import {execFileSync, spawnSync} from 'child_process';
+import {readFileSync} from 'fs';
 import path from 'path';
 import {fileURLToPath} from 'url';
 
@@ -87,11 +89,51 @@ function runCSpell(files) {
 		column: issue.col,
 		word: issue.text,
 		suggestions: issue.suggestions?.slice(0, 3) ?? [],
-		text: issue.line.text.replace(/\r?\n$/, ''),
-		context: issue.context?.text.trim() ?? issue.line.text.trim()
+		text: issue.line.text.replace(/\r?\n$/, '')
 	}));
 }
 
+// Entries in the dictionary files must be sorted, so duplicates stand out and
+// merges are clean. Sorted the way CSpell matches: case- and accent-insensitive.
+// Blank lines and comments start a new sorted run, so the lists can be sectioned.
+const DICTIONARY_FILES = ['cspell-allow-list.txt', 'cspell-block-list.txt'];
+const collator = new Intl.Collator('en', {sensitivity: 'base'});
+
+function unsortedDictionaryEntries(files) {
+	const findings = [];
+	for (const file of files.filter(file => DICTIONARY_FILES.includes(file))) {
+		let run = []; // [{word, line}] of the current sorted run
+		readFileSync(file, 'utf8')
+			.split('\n')
+			.forEach((text, index) => {
+				if (/^\s*(#|$)/.test(text)) {
+					run = [];
+					return;
+				}
+				const word = text.replace(/\s*#.*/, '').trim();
+				const line = index + 1;
+				const belongsBefore = run.find(
+					entry => collator.compare(word, entry.word) < 0
+				);
+				if (belongsBefore) {
+					findings.push({
+						file,
+						line,
+						column: 1,
+						word,
+						suggestions: [],
+						text,
+						message: `\`${word}\` is out of alphabetical order: move it above \`${belongsBefore.word}\``,
+						relatedLine: belongsBefore.line // rendered as a link after the message
+					});
+				} else {
+					run.push({word, line});
+				}
+			});
+	}
+	return findings;
+}
+
 function findingsOnAddedLines(ranges, issues) {
 	return issues.filter(issue =>
 		(ranges.get(issue.file) ?? []).some(
@@ -102,16 +144,17 @@ function findingsOnAddedLines(ranges, issues) {
 
 function formatText(findings) {
 	if (findings.length === 0) {
-		return 'No spelling errors found in added lines.\n';
+		return 'No issues found in added lines.\n';
 	}
 
 	const lines = [
-		`Found ${findings.length} spelling error(s) in added lines:`
+		`Found ${findings.length} issue(s) in added lines:`
 	];
-	for (const finding of findings) {
-		lines.push(
-			`${finding.file}:${finding.line}:${finding.column} - Unknown word (${finding.word})`
-		);
+	for (const {file, line, column, word, message, relatedLine} of findings) {
+		const detail = message
+			? `${message}${relatedLine ? ` on line ${relatedLine}` : ''}`
+			: `Unknown word (${word})`;
+		lines.push(`${file}:${line}:${column} - ${detail}`);
 	}
 	return lines.join('\n') + '\n';
 }
@@ -125,10 +168,11 @@ async function main() {
 	}
 
 	const ranges = addedLineRanges(BASE);
-	const findings = findingsOnAddedLines(
-		ranges,
-		runCSpell([...ranges.keys()])
-	);
+	const files = [...ranges.keys()];
+	const findings = findingsOnAddedLines(ranges, [
+		...runCSpell(files),
+		...unsortedDictionaryEntries(files)
+	]);
 	process.stdout.write(
 		FORMAT === 'json'
 			? JSON.stringify(findings, null, '\t') + '\n'
diff --git a/dev/post-spelling-review.mjs b/dev/post-spelling-review.mjs
index 38d18a88e..94eacb399 100644
--- a/dev/post-spelling-review.mjs
+++ b/dev/post-spelling-review.mjs
@@ -100,10 +100,35 @@ function groupByFile(findings) {
 	return grouped;
 }
 
+// Source view (?plain=1, so Markdown is not rendered) with the word highlighted
+function sourceLink(finding) {
+	const {file, line, column, word} = finding;
+	const end = column + word.length;
+	return `https://github.com/${REPOSITORY}/blob/${HEAD_REF}/${file}?plain=1#L${line}C${column}-L${line}C${end}`;
+}
+
+// A non-spelling finding's own message, e.g. an unsorted dictionary entry,
+// pointing at its `relatedLine` when it has one
+function messageText(finding) {
+	const {file, message, relatedLine} = finding;
+	if (!relatedLine) return `${message}.`;
+	const url = `https://github.com/${REPOSITORY}/blob/${HEAD_REF}/${file}?plain=1#L${relatedLine}`;
+	return `${message} on [line ${relatedLine}](${url}).`;
+}
+
+// `word` → `suggestion`, or the finding's own message
+function summaryItem(finding) {
+	if (finding.message) {
+		return messageText(finding);
+	}
+	const suggestion = bestSuggestion(finding);
+	return `\`${finding.word}\`${suggestion ? ` → \`${suggestion}\`` : ''}`;
+}
+
 function summaryBody(findings) {
 	const lines = [
 		SUMMARY_MARKER,
-		`### ⚠️ CSpell found ${findings.length} spelling error(s) in this PR`,
+		`### ⚠️ Spell check found ${findings.length} issue(s) in this PR`,
 		'',
 		'Only findings on lines added by this PR are shown.',
 		...(findings.length > MAX_INLINE_COMMENTS
@@ -115,10 +140,11 @@ function summaryBody(findings) {
 	];
 	for (const [file, fileFindings] of groupByFile(findings)) {
 		lines.push(`**\`${file}\`**`);
-		for (const {line, column, word, context} of fileFindings) {
-			const excerpt = context.replaceAll('`', "'").slice(0, 160);
+		for (const finding of fileFindings) {
+			const {line, column} = finding;
 			lines.push(
-				`- line ${line}, column ${column}: \`${word}\` — \`${excerpt}\``
+				`- [line ${line}, column ${column}](${sourceLink(finding)})`,
+				`  - ${summaryItem(finding)}`
 			);
 		}
 		lines.push('');
@@ -207,13 +233,15 @@ function suggestionBlock(finding) {
 }
 
 function inlineBody(finding) {
-	return [
-		`${INLINE_MARKER} ${finding.word} -->`,
-		`\`${finding.word}\` is not in the dictionary.`,
-		'',
-		...suggestionBlock(finding),
-		`Please correct the spelling, or add the word to ${ALLOW_LIST_LINK} if it is correct.`
-	].join('\n');
+	const explanation = finding.message
+		? [messageText(finding)]
+		: [
+				`\`${finding.word}\` is not in the dictionary.`,
+				'',
+				...suggestionBlock(finding),
+				`Please correct the spelling, or add the word to ${ALLOW_LIST_LINK} if it is correct.`
+			];
+	return [`${INLINE_MARKER} ${finding.word} -->`, ...explanation].join('\n');
 }
 
 async function syncInlineComments(findings) {