diff --git a/infra/flue-review/.flue/lib/diff-budget.ts b/infra/flue-review/.flue/lib/diff-budget.ts
new file mode 100644
index 0000000000..526b43b172
--- /dev/null
+++ b/infra/flue-review/.flue/lib/diff-budget.ts
@@ -0,0 +1,74 @@
+// Size budget for the staged PR diff. The review agent starts by reading the
+// whole diff file, so an oversized diff (generated types, lockfiles, large
+// catalogs) lands in the model context verbatim and kills the model call.
+// Oversized per-file sections are elided down to their headers with a note;
+// the agent reads those files from the checkout instead.
+
+const DEFAULT_PER_FILE_BYTES = 48 * 1024;
+const DEFAULT_TOTAL_BYTES = 384 * 1024;
+
+export interface DiffBudget {
+	readonly perFileBytes?: number;
+	readonly totalBytes?: number;
+}
+
+interface Section {
+	text: string;
+	elided: boolean;
+}
+
+export function elideLargeDiffSections(diff: string, budget: DiffBudget = {}): string {
+	const perFileBytes = budget.perFileBytes ?? DEFAULT_PER_FILE_BYTES;
+	const totalBytes = budget.totalBytes ?? DEFAULT_TOTAL_BYTES;
+	if (diff.length <= Math.min(perFileBytes, totalBytes)) return diff;
+
+	const sections = splitSections(diff);
+	for (const section of sections) {
+		if (!section.elided && section.text.length > perFileBytes) elide(section);
+	}
+	// Still over the total budget: elide the largest remaining sections until
+	// under it (or nothing left to elide).
+	let total = sections.reduce((n, s) => n + s.text.length, 0);
+	while (total > totalBytes) {
+		const next = sections
+			.filter((s) => !s.elided)
+			.toSorted((a, b) => b.text.length - a.text.length)[0];
+		if (!next) break;
+		total -= next.text.length;
+		elide(next);
+		total += next.text.length;
+	}
+	return sections.map((s) => s.text).join("");
+}
+
+function splitSections(diff: string): Section[] {
+	const starts: number[] = [];
+	const re = /^diff --git /gm;
+	for (let m = re.exec(diff); m; m = re.exec(diff)) starts.push(m.index);
+	if (starts.length === 0) return [{ text: diff, elided: false }];
+	const sections: Section[] = [];
+	if (starts[0] !== 0) sections.push({ text: diff.slice(0, starts[0]), elided: true });
+	for (let i = 0; i < starts.length; i++) {
+		const end = i + 1 < starts.length ? starts[i + 1] : diff.length;
+		sections.push({ text: diff.slice(starts[i], end), elided: false });
+	}
+	return sections;
+}
+
+function elide(section: Section): void {
+	// Mark unconditionally: a section this function cannot reduce must still
+	// leave the total-budget loop's candidate pool, or the loop never shrinks.
+	section.elided = true;
+	const lines = section.text.split("\n");
+	// Keep the file header: everything up to and including the `+++` line (or
+	// the whole header for binary/rename-only sections with no hunks).
+	let headerEnd = lines.findIndex((line) => line.startsWith("+++ "));
+	if (headerEnd === -1) headerEnd = lines.findIndex((line) => line.startsWith("@@ ")) - 1;
+	if (headerEnd < 0) return;
+	const body = lines.length - (headerEnd + 1);
+	section.text = [
+		...lines.slice(0, headerEnd + 1),
+		`(diff content elided: ${body} lines over the size budget -- read this file from the checkout instead)`,
+		"",
+	].join("\n");
+}
diff --git a/infra/flue-review/.flue/workflows/review.ts b/infra/flue-review/.flue/workflows/review.ts
index 651f4beec1..dd4f631acc 100644
--- a/infra/flue-review/.flue/workflows/review.ts
+++ b/infra/flue-review/.flue/workflows/review.ts
@@ -27,6 +27,7 @@ import { getCloudflareContext, getDurableObjectIdentity } from "@flue/runtime/cl
 import * as v from "valibot";
 
 import { withCapacityRetry } from "../lib/capacity.js";
+import { elideLargeDiffSections } from "../lib/diff-budget.js";
 import {
 	readAppCreds,
 	mintInstallationToken,
@@ -358,7 +359,7 @@ async function run(context: ActionContext<typeof reviewPayloadSchema>): Promise<
 			payload.baseSha,
 			payload.headSha,
 		);
-		await context.harness.fs.writeFile(DIFF_PATH, diff);
+		await context.harness.fs.writeFile(DIFF_PATH, elideLargeDiffSections(diff));
 
 		stage = "model_review";
 		if (
diff --git a/infra/flue-review/test/diff-budget.test.ts b/infra/flue-review/test/diff-budget.test.ts
new file mode 100644
index 0000000000..511e65805f
--- /dev/null
+++ b/infra/flue-review/test/diff-budget.test.ts
@@ -0,0 +1,71 @@
+import { describe, expect, it } from "vitest";
+
+import { elideLargeDiffSections } from "../.flue/lib/diff-budget.js";
+
+function fileSection(path: string, lines: number, line = "+const x = 1;"): string {
+	return [
+		`diff --git a/${path} b/${path}`,
+		"index 0000000..1111111 100644",
+		`--- a/${path}`,
+		`+++ b/${path}`,
+		"@@ -0,0 +1 @@",
+		...Array.from({ length: lines }).fill(line).map(String),
+		"",
+	].join("\n");
+}
+
+describe("elideLargeDiffSections", () => {
+	it("returns a small diff unchanged", () => {
+		const diff = fileSection("src/a.ts", 10) + fileSection("src/b.ts", 20);
+		expect(elideLargeDiffSections(diff)).toBe(diff);
+	});
+
+	it("elides a section over the per-file budget, keeping its header", () => {
+		const big = fileSection("types.d.ts", 2_000);
+		const small = fileSection("src/a.ts", 5);
+		const out = elideLargeDiffSections(small + big, { perFileBytes: 1_000 });
+		expect(out).toContain(small);
+		expect(out).toContain("diff --git a/types.d.ts b/types.d.ts");
+		expect(out).toContain("+++ b/types.d.ts");
+		expect(out).toMatch(/diff content elided: \d+ lines/);
+		expect(out).not.toContain("+const x = 1;\n+const x = 1;\n".repeat(50));
+	});
+
+	it("elides largest sections first until under the total budget", () => {
+		const a = fileSection("a.ts", 30);
+		const b = fileSection("b.ts", 60);
+		const c = fileSection("c.ts", 10);
+		const out = elideLargeDiffSections(a + b + c, {
+			perFileBytes: 10_000,
+			totalBytes: a.length + c.length + 400,
+		});
+		expect(out).toContain("diff content elided");
+		expect(out).toContain(a);
+		expect(out).toContain(c);
+		expect(out).not.toContain(b);
+	});
+
+	it("skips an unreducible headerless section instead of looping on it", () => {
+		const headerless = `diff --git a/blob.bin b/blob.bin\nBinary files differ\n${"x\n".repeat(1_000)}`;
+		const small = fileSection("src/a.ts", 5);
+		const out = elideLargeDiffSections(small + headerless, {
+			perFileBytes: 500,
+			totalBytes: 600,
+		});
+		expect(out).toContain("Binary files differ");
+		expect(out).toContain("+++ b/src/a.ts");
+	});
+
+	it("leaves a header-only section (no hunks) alone", () => {
+		const rename = [
+			"diff --git a/old.ts b/new.ts",
+			"similarity index 100%",
+			"rename from old.ts",
+			"rename to new.ts",
+			"",
+		].join("\n");
+		const filler = fileSection("big.ts", 2_000);
+		const out = elideLargeDiffSections(rename + filler, { perFileBytes: 1_000, totalBytes: 1_500 });
+		expect(out).toContain("rename from old.ts");
+	});
+});