pi-agent-browser-native 0.2.44 → 0.2.45

package/extensions/agent-browser/lib/results/presentation.ts

@@ -66,6 +66,10 @@ function mergeNextActions(...groups: Array<AgentBrowserNextAction[] | undefined>
 	return merged.length > 0 ? merged : undefined;
 }
 
+function shouldAddAnnotatedScreenshotGuidance(commandInfo: CommandInfo, args: string[] | undefined): boolean {
+	return commandInfo.command === "screenshot" && (args?.includes("--annotate") ?? false);
+}
+
 export async function buildToolPresentation(options: {
 	artifactManifest?: SessionArtifactManifest;
 	args?: string[];
@@ -103,7 +107,7 @@ export async function buildToolPresentation(options: {
 
 	const data = enrichStreamStatusData(commandInfo, envelope?.data);
 	const presentationData = redactPresentationData(commandInfo, data);
-	const artifacts = await extractFileArtifacts({ artifactRequest, commandInfo: presentationCommandInfo, cwd, data, sessionName });
+	const artifacts = await extractFileArtifacts({ artifactManifest, artifactRequest, commandInfo: presentationCommandInfo, cwd, data, sessionName });
 	const artifactVerification = buildArtifactVerificationSummary(artifacts);
 	const artifactSummary = formatArtifactSummary(artifacts);
 	const summary = artifactSummary ?? formatPresentationSummary(commandInfo, data, compiledSemanticAction);
@@ -151,6 +155,11 @@ export async function buildToolPresentation(options: {
 		}
 	}
 
+	if (shouldAddAnnotatedScreenshotGuidance(commandInfo, args) && presentation.content[0]?.type === "text") {
+		const guidance = "Annotated screenshot note: dense pages can produce overlapping labels. If the labels are noisy, capture a scoped element screenshot, take a non-annotated screenshot, or use snapshot -i high-value refs as the machine-readable map.";
+		presentation.content[0] = { ...presentation.content[0], text: `${presentation.content[0].text}\n\n${guidance}` };
+	}
+
 	const imagePath = artifactRequest?.absolutePath ?? extractImagePath(commandInfo, cwd, data);
 	const presentationWithImage = imagePath ? await attachInlineImage(presentation, imagePath) : presentation;
 	const compactedPresentation = await compactLargePresentationOutput({

package/extensions/agent-browser/lib/results/snapshot-high-value-controls.ts

@@ -17,6 +17,7 @@ const SNAPSHOT_HIGH_VALUE_CONTROL_ROLES = new Set([
 	"button",
 	"checkbox",
 	"combobox",
+	"link",
 	"menuitem",
 	"option",
 	"radio",
@@ -30,11 +31,12 @@ const SNAPSHOT_HIGH_VALUE_CONTROL_ROLE_PRIORITY: Record<string, number> = {
 	textbox: 1,
 	combobox: 2,
 	button: 3,
-	tab: 4,
-	checkbox: 5,
-	radio: 6,
-	option: 7,
-	menuitem: 8,
+	link: 4,
+	tab: 5,
+	checkbox: 6,
+	radio: 7,
+	option: 8,
+	menuitem: 9,
 };
 
 const SNAPSHOT_SURFACE_CONTROL_NAME_PATTERNS = [
@@ -46,6 +48,10 @@ const SNAPSHOT_PRIMARY_ACTION_BUTTON_NAME_PATTERNS = [
 	/^(?:add|apply|ask|confirm|connect|continue|create|launch|new|open|refresh|retry|run|save|search|send|start|submit)\b/i,
 ];
 
+const SNAPSHOT_HIGH_VALUE_LINK_NAME_PATTERNS = [
+	/^[a-z0-9_.-]+\/[a-z0-9_.-]+$/i,
+];
+
 function getHighValueControlRole(entry: SnapshotRefEntry): string {
 	return entry.isEditable === true && (entry.role === "unknown" || entry.role === "generic") ? "textbox" : entry.role;
 }
@@ -129,9 +135,14 @@ const SNAPSHOT_HIGH_VALUE_CONTROL_CATEGORY_RULES: readonly HighValueControlCateg
 	},
 ] as const;
 
+function isHighValueLinkRef(entry: SnapshotRefEntry): boolean {
+	return entry.name.length > 0 && SNAPSHOT_HIGH_VALUE_LINK_NAME_PATTERNS.some((pattern) => pattern.test(entry.name));
+}
+
 export function isHighValueControlEntry(entry: SnapshotRefEntry): boolean {
 	const role = getHighValueControlRole(entry);
 	if (!SNAPSHOT_HIGH_VALUE_CONTROL_ROLES.has(role)) return false;
+	if (role === "link") return isHighValueLinkRef(entry);
 	if (entry.isEditable === false && (role === "searchbox" || role === "textbox" || role === "combobox")) return false;
 	return entry.name.length > 0 || isEditableControlRef(entry);
 }

package/extensions/agent-browser/lib/results/snapshot.ts

@@ -221,6 +221,7 @@ export async function buildSnapshotPresentation(
 		...(roleCountsText ? [`Top roles: ${roleCountsText}`] : []),
 		"",
 		"Compact snapshot view.",
+		"Viewport note: compact snapshots are DOM/signal-prioritized, not guaranteed to start with the currently scrolled viewport; use the full raw snapshot, a screenshot, or listed high-value refs when viewport context matters.",
 	];
 
 	if (fallbackPreview) {
@@ -294,6 +295,7 @@ export async function buildSnapshotPresentation(
 			fullOutputPath,
 			origin,
 			previewMode: fallbackPreview ? "outline" : "structured",
+			viewportOrdering: "dom-signal-prioritized",
 			spillError: spillErrorText,
 			previewRefIds: [...previewRefIds],
 			highValueControlRefIds: visibleHighValueControlEntries.map((entry) => entry.id),

package/extensions/agent-browser/lib/runtime.ts

@@ -619,6 +619,15 @@ export function createFreshSessionName(baseSessionName: string, ephemeralSeed: s
 	return `${baseSessionName}-fresh-${suffix}`;
 }
 
+function getSingleKeyCommandValidationError(args: string[]): string | undefined {
+	const { commandInfo, commandTokens } = parseArgvDescriptor(args);
+	const command = commandInfo.command;
+	if (command !== "press" && command !== "key" && command !== "keydown" && command !== "keyup") return undefined;
+	if (commandTokens.length === 2) return undefined;
+	const label = command === "key" ? "key/press" : command;
+	return `agent-browser ${label} accepts exactly one key argument. Do not pass a selector or ref to ${label}; focus or click the target first, then run ${command} <key> (for example: focus @e1, then press Enter).`;
+}
+
 export function validateToolArgs(args: string[]): string | undefined {
 	if (args.length === 0) {
 		return "`args` must contain at least one agent-browser command token.";
@@ -634,7 +643,7 @@ export function validateToolArgs(args: string[]): string | undefined {
 		return "Do not pass `--session-mode` in args. Use the top-level agent_browser `sessionMode` field instead, for example { args: [\"--profile\", \"Default\", \"open\", \"https://example.com\"], sessionMode: \"fresh\" }.";
 	}
 
-	return undefined;
+	return getSingleKeyCommandValidationError(args);
 }
 
 function getInvalidValueFlagDetails(args: string[]): InvalidValueFlagDetails | undefined {

package/extensions/agent-browser/lib/session-page-state.ts

@@ -8,6 +8,9 @@
 
 import { isCloseCommand, isReadOnlyDiagnosticSessionTargetCommand } from "./command-taxonomy.js";
 import { isRecord } from "./parsing.js";
+import { getEditableRefEvidence } from "./results/editable-ref-evidence.js";
+import { enrichSnapshotRefEntries, getSnapshotRefEntries } from "./results/snapshot-refs.js";
+import { parseSnapshotLines } from "./results/snapshot-segments.js";
 
 export interface SessionTabTarget {
 	title?: string;
@@ -21,7 +24,7 @@ interface OrderedSessionTabTarget {
 
 export interface SessionRefSnapshot {
 	refIds: string[];
-	refs?: Record<string, { name: string; role: string }>;
+	refs?: Record<string, { isContentEditable?: boolean; isEditable?: boolean; name: string; role: string }>;
 	target?: SessionTabTarget;
 }
 
@@ -230,11 +233,15 @@ function getRestoredSessionTabTarget(details: Record<string, unknown>, command:
 	return storedTarget;
 }
 
-function extractRefSnapshotRefs(data: unknown): Record<string, { name: string; role: string }> | undefined {
+function extractRefSnapshotRefs(data: unknown): Record<string, { isContentEditable?: boolean; isEditable?: boolean; name: string; role: string }> | undefined {
 	if (!isRecord(data) || !isRecord(data.refs)) return undefined;
-	const refs = Object.fromEntries(Object.entries(data.refs).flatMap(([refId, entry]) => {
-		if (!/^e\d+$/.test(refId) || !isRecord(entry) || typeof entry.name !== "string" || typeof entry.role !== "string") return [];
-		return [[refId, { name: entry.name, role: entry.role }] as const];
+	const snapshotLines = typeof data.snapshot === "string" ? parseSnapshotLines(data.snapshot) : [];
+	const lineByRef = new Map(snapshotLines.flatMap((line) => line.ref ? [[line.ref, line.raw] as const] : []));
+	const entries = enrichSnapshotRefEntries(getSnapshotRefEntries(data), snapshotLines);
+	const refs = Object.fromEntries(entries.flatMap((entry) => {
+		if (!/^e\d+$/.test(entry.id) || entry.role.length === 0) return [];
+		const isContentEditable = getEditableRefEvidence({ ref: entry.refData, text: lineByRef.get(entry.id) });
+		return [[entry.id, { ...(isContentEditable === true ? { isContentEditable: true } : {}), ...(entry.isEditable !== undefined ? { isEditable: entry.isEditable } : {}), name: entry.name, role: entry.role }] as const];
 	}));
 	return Object.keys(refs).length > 0 ? refs : undefined;
 }
@@ -310,7 +317,9 @@ function getRestoredRefSnapshot(details: Record<string, unknown>): SessionRefSna
 		? Object.fromEntries(refIds.flatMap((refId) => {
 			const entry = refRecord[refId];
 			if (!isRecord(entry) || typeof entry.name !== "string" || typeof entry.role !== "string") return [];
-			return [[refId, { name: entry.name, role: entry.role }] as const];
+			const isContentEditable = typeof entry.isContentEditable === "boolean" ? entry.isContentEditable : undefined;
+			const isEditable = typeof entry.isEditable === "boolean" ? entry.isEditable : undefined;
+			return [[refId, { ...(isContentEditable !== undefined ? { isContentEditable } : {}), ...(isEditable !== undefined ? { isEditable } : {}), name: entry.name, role: entry.role }] as const];
 		}))
 		: undefined;
 	return {

package/extensions/agent-browser/lib/web-search.ts

@@ -653,7 +653,7 @@ export function createAgentBrowserWebSearchTool(configState: AgentBrowserConfigS
 		promptSnippet: "Search the live web with Exa or Brave for current or external information.",
 		promptGuidelines: [
 			"Use agent_browser_web_search when live web search would help answer the task, find current external information, or discover candidate URLs for agent_browser.",
-			"The tool chooses Exa or Brave from configured keys; when both are available, Exa is preferred by default unless webSearch.preferredProvider says otherwise. Use provider only when the user/config calls for a specific provider.",
+			"agent_browser_web_search chooses Exa or Brave from configured keys; when both are available, Exa is preferred by default unless webSearch.preferredProvider says otherwise. Use provider only when the user/config calls for a specific provider.",
 			"Prefer agent_browser_web_search over opening a search engine results page with agent_browser when a quick result list is enough; use agent_browser for interaction, DOM, screenshots, or auth.",
 			"Do not issue parallel or repeated agent_browser_web_search calls; use one high-signal query, inspect the results, then only run a focused follow-up if needed. If the provider returns HTTP 429, stop searching and tell the user the API plan/rate limit needs time or a plan change.",
 			"After using agent_browser_web_search, cite result URLs in the final answer when web evidence informed the answer.",

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "pi-agent-browser-native",
-  "version": "0.2.44",
+  "version": "0.2.45",
   "description": "pi extension that exposes agent-browser as a native tool for browser automation",
   "type": "module",
   "author": "Mitch Fultz (https://github.com/fitchmultz)",
@@ -80,7 +80,7 @@
     "check:platform-smoke": "node --check platform-smoke.config.mjs && node --check scripts/platform-smoke.mjs && node --check scripts/platform-smoke/doctor.mjs && node --check scripts/platform-smoke/crabbox-runner.mjs && node --check scripts/platform-smoke/targets.mjs && node --check scripts/platform-smoke/artifacts.mjs && tsx --test test/platform-smoke.test.ts",
     "smoke:platform": "node scripts/platform-smoke.mjs",
     "smoke:platform:doctor": "node scripts/platform-smoke.mjs doctor",
-    "smoke:platform:ubuntu-image": "docker build -t pi-agent-browser-native-platform:node24-agent-browser0.27.1 --build-arg AGENT_BROWSER_VERSION=0.27.1 -f scripts/platform-smoke/linux-image/Dockerfile .",
+    "smoke:platform:ubuntu-image": "node scripts/platform-smoke/build-ubuntu-image.mjs",
     "smoke:platform:macos": "node scripts/platform-smoke.mjs run --target macos",
     "smoke:platform:ubuntu": "node scripts/platform-smoke.mjs run --target ubuntu",
     "smoke:platform:windows-native": "node scripts/platform-smoke.mjs run --target windows-native",

package/platform-smoke.config.mjs

@@ -3,6 +3,9 @@
 
 import { CAPABILITY_BASELINE } from "./scripts/agent-browser-capability-baseline.mjs";
 
+export const PLATFORM_SMOKE_AGENT_BROWSER_VERSION = CAPABILITY_BASELINE.targetVersion;
+export const PLATFORM_SMOKE_UBUNTU_IMAGE = `pi-agent-browser-native-platform:node24-agent-browser${PLATFORM_SMOKE_AGENT_BROWSER_VERSION}`;
+
 export default {
 	packageName: "pi-agent-browser-native",
 	artifactRoot: ".artifacts/platform-smoke",
@@ -17,11 +20,11 @@ export default {
 		host: "localhost",
 		port: 22,
 	},
-	ubuntuContainerImage: "pi-agent-browser-native-platform:node24-agent-browser0.27.1",
+	ubuntuContainerImage: PLATFORM_SMOKE_UBUNTU_IMAGE,
 	windowsParallels: {
 		sourceVm: "pi-extension-windows-template",
 		snapshot: "crabbox-ready",
 	},
 	nodeValidationMajor: 22,
-	agentBrowserVersion: CAPABILITY_BASELINE.targetVersion,
+	agentBrowserVersion: PLATFORM_SMOKE_AGENT_BROWSER_VERSION,
 };

package/scripts/platform-smoke/build-ubuntu-image.mjs

@@ -0,0 +1,25 @@
+#!/usr/bin/env node
+import { spawnSync } from "node:child_process";
+
+import { CAPABILITY_BASELINE } from "../agent-browser-capability-baseline.mjs";
+
+const version = CAPABILITY_BASELINE.targetVersion;
+const image = `pi-agent-browser-native-platform:node24-agent-browser${version}`;
+const args = [
+	"build",
+	"-t",
+	image,
+	"--build-arg",
+	`AGENT_BROWSER_VERSION=${version}`,
+	"-f",
+	"scripts/platform-smoke/linux-image/Dockerfile",
+	".",
+];
+
+console.log(`Building ${image}`);
+const result = spawnSync("docker", args, { stdio: "inherit" });
+if (result.error) {
+	console.error(result.error.message);
+	process.exit(1);
+}
+process.exit(result.status ?? 1);

package/scripts/platform-smoke/crabbox-runner.mjs

@@ -2,6 +2,10 @@
 
 import { spawn } from "node:child_process";
 
+import { CAPABILITY_BASELINE } from "../agent-browser-capability-baseline.mjs";
+
+const DEFAULT_UBUNTU_IMAGE = `pi-agent-browser-native-platform:node24-agent-browser${CAPABILITY_BASELINE.targetVersion}`;
+
 function env(name) {
 	return process.env[name] ?? "";
 }
@@ -38,7 +42,7 @@ export function describeTarget(targetName, config = {}) {
 			};
 		}
 		case "ubuntu": {
-			const image = env("PLATFORM_SMOKE_UBUNTU_IMAGE") || config.ubuntuContainerImage || "pi-agent-browser-native-platform:node24-agent-browser0.27.1";
+			const image = env("PLATFORM_SMOKE_UBUNTU_IMAGE") || config.ubuntuContainerImage || DEFAULT_UBUNTU_IMAGE;
 			return {
 				provider: "local-container",
 				crabboxTarget: "linux",

package/scripts/platform-smoke/doctor.mjs

@@ -4,6 +4,10 @@ import { execFileSync, execSync } from "node:child_process";
 import { accessSync, constants, mkdirSync, unlinkSync, writeFileSync } from "node:fs";
 import { resolve } from "node:path";
 
+import { CAPABILITY_BASELINE } from "../agent-browser-capability-baseline.mjs";
+
+const DEFAULT_UBUNTU_IMAGE = `pi-agent-browser-native-platform:node24-agent-browser${CAPABILITY_BASELINE.targetVersion}`;
+
 function env(name) {
 	return process.env[name] ?? "";
 }
@@ -220,7 +224,7 @@ export async function runDoctor(config) {
 	console.log("\n── Crabbox providers ──");
 	if (cboxPath) {
 		checkRequiredProviders(cbox, failures);
-		const ubuntuImage = env("PLATFORM_SMOKE_UBUNTU_IMAGE") || config?.ubuntuContainerImage || "pi-agent-browser-native-platform:node24-agent-browser0.27.1";
+		const ubuntuImage = env("PLATFORM_SMOKE_UBUNTU_IMAGE") || config?.ubuntuContainerImage || DEFAULT_UBUNTU_IMAGE;
 		checkCrabboxProvider(cbox, ["--provider", "local-container", "--local-container-image", ubuntuImage], "ubuntu local-container", failures);
 		const macUser = env("PLATFORM_SMOKE_MAC_USER") || env("USER");
 		const macHost = env("PLATFORM_SMOKE_MAC_HOST") || config?.macos?.host || "localhost";
@@ -233,7 +237,7 @@ export async function runDoctor(config) {
 	const dockerVersion = shell("docker info --format '{{.ServerVersion}}'");
 	if (dockerVersion) ok(`Docker ${dockerVersion}`);
 	else fail("Docker is not available or not running", failures);
-	const ubuntuImage = env("PLATFORM_SMOKE_UBUNTU_IMAGE") || config?.ubuntuContainerImage || "pi-agent-browser-native-platform:node24-agent-browser0.27.1";
+	const ubuntuImage = env("PLATFORM_SMOKE_UBUNTU_IMAGE") || config?.ubuntuContainerImage || DEFAULT_UBUNTU_IMAGE;
 	ok(`Ubuntu image: ${ubuntuImage}`);
 
 	console.log("\n── macOS SSH ──");

package/scripts/platform-smoke/linux-image/Dockerfile

@@ -1,12 +1,10 @@
 # Local Crabbox Ubuntu/Linux target image for pi-agent-browser-native platform smoke.
-# Build with:
-#   docker build -t pi-agent-browser-native-platform:node24-agent-browser0.27.1 \
-#     --build-arg AGENT_BROWSER_VERSION=0.27.1 \
-#     -f scripts/platform-smoke/linux-image/Dockerfile .
+# Build with npm run smoke:platform:ubuntu-image so the agent-browser version
+# comes from scripts/agent-browser-capability-baseline.mjs.
 
 FROM node:24-bookworm
 
-ARG AGENT_BROWSER_VERSION=0.27.1
+ARG AGENT_BROWSER_VERSION
 
 USER root
 RUN apt-get update \

package/scripts/platform-smoke/targets.mjs

@@ -15,6 +15,7 @@ import {
 	writeManifest,
 	writeSummary,
 } from "./artifacts.mjs";
+import { CAPABILITY_BASELINE } from "../agent-browser-capability-baseline.mjs";
 import { cleanupStaleTargetState, crabboxBin, describeTarget, runOnLease, stopLease, warmupLease } from "./crabbox-runner.mjs";
 
 export function platformFor(targetName) {
@@ -300,7 +301,7 @@ export function buildPlatformBuildCommand(targetName, packageName = "pi-agent-br
 	return lines.join("\n");
 }
 
-export function buildBrowserDogfoodCommand(targetName, agentBrowserVersion = "0.27.1") {
+export function buildBrowserDogfoodCommand(targetName, agentBrowserVersion = CAPABILITY_BASELINE.targetVersion) {
 	if (platformFor(targetName) === "powershell") {
 		return `powershell.exe -NoLogo -NoProfile -ExecutionPolicy Bypass -File .\\scripts\\platform-smoke\\browser-dogfood-windows.ps1 -AgentBrowserVersion ${psSingleQuote(agentBrowserVersion)}`;
 	}

package/extensions/agent-browser/lib/orchestration/browser-run/browser-action-model.ts

@@ -1,154 +0,0 @@
-/**
- * Purpose: Normalize planned browser argv into a small action model for prompt-derived guards.
- * Responsibilities: Map command tokens and batch stdin steps to click-like and keyboard-submit actions with target labels.
- * Scope: Best-effort finalizing-action detection only; does not model eval, generic fill/type, or non-Enter keyboard flows.
- */
-
-import type { SessionRefSnapshot } from "../../session-page-state.js";
-import { parseValidBatchStepEntries } from "../batch-stdin.js";
-
-const FINAL_ACTION_PATTERN = /\b(?:finish|place\s+(?:the\s+)?order|submit\s+(?:the\s+)?order|complete\s+(?:the\s+)?order|confirm\s+(?:the\s+)?order|purchase|buy\s+now|pay\s+now|finali[sz]e|submit\s+payment|checkout\s+complete)\b/i;
-
-const CLICK_LIKE_COMMANDS = new Set(["click", "dblclick", "tap"]);
-const FIND_CLICK_ACTIONS = new Set(["click", "dblclick", "tap"]);
-const KEYBOARD_SUBMIT_KEYS = new Set(["enter", "return"]);
-
-export type BrowserFinalizingActionKind = "click-like" | "keyboard-submit";
-
-export interface BrowserFinalizingAction {
-	command: string[];
-	kind: BrowserFinalizingActionKind;
-	stepIndex?: number;
-	targetLabel?: string;
-}
-
-export const STOP_BOUNDARY_GUARD_SCOPE = {
-	covered: [
-		"standalone click, dblclick, and tap",
-		"find … click|dblclick|tap",
-		"batch steps with the click-like shapes above",
-		"press <key> and key <key> when key is Enter or Return",
-	],
-	excluded: [
-		"eval --stdin and other scripted activation",
-		"fill, type, select, drag, and upload without an explicit click-like command",
-		"keyboard type/inserttext and keyboard shortcuts other than Enter/Return",
-		"semanticAction and job/qa compiled plans unless their batch stdin contains a covered step",
-	],
-} as const;
-
-function normalizeTargetText(value: string): string {
-	return value
-		.replace(/[_-]+/g, " ")
-		.replace(/[\[\]{}()#.'\"=:/]+/g, " ")
-		.replace(/\s+/g, " ")
-		.trim();
-}
-
-export function matchesFinalActionLabel(value: string | undefined): boolean {
-	return value !== undefined && FINAL_ACTION_PATTERN.test(normalizeTargetText(value));
-}
-
-function parseRefId(value: string | undefined): string | undefined {
-	if (!value) return undefined;
-	const trimmed = value.trim();
-	const candidate = trimmed.startsWith("@") ? trimmed.slice(1) : trimmed.startsWith("ref=") ? trimmed.slice(4) : trimmed;
-	return /^e\d+$/.test(candidate) ? candidate : undefined;
-}
-
-function getRefTargetLabel(refSnapshot: SessionRefSnapshot | undefined, refId: string | undefined): string | undefined {
-	if (!refId) return undefined;
-	const ref = refSnapshot?.refs?.[refId];
-	return ref ? [ref.role, ref.name].filter(Boolean).join(" ") : undefined;
-}
-
-function getFlagValue(tokens: string[], flag: string): string | undefined {
-	for (const [index, token] of tokens.entries()) {
-		if (token === flag) return tokens[index + 1];
-		if (token.startsWith(`${flag}=`)) return token.slice(flag.length + 1);
-	}
-	return undefined;
-}
-
-function getClickLikeTargetLabel(command: string[], refSnapshot: SessionRefSnapshot | undefined): string | undefined {
-	const target = command[1];
-	return getRefTargetLabel(refSnapshot, parseRefId(target)) ?? target;
-}
-
-function getFindClickTargetLabel(command: string[]): string | undefined {
-	if (command[0] !== "find") return undefined;
-	const actionIndex = command.findIndex((token, index) => index >= 3 && FIND_CLICK_ACTIONS.has(token));
-	if (actionIndex === -1) return undefined;
-	return getFlagValue(command, "--name") ?? command[2];
-}
-
-function getKeyboardSubmitKey(command: string[]): string | undefined {
-	const commandName = command[0];
-	if (commandName === "press" || commandName === "key") return command[1];
-	return undefined;
-}
-
-function collectActionsFromCommand(command: string[], refSnapshot: SessionRefSnapshot | undefined, stepIndex?: number): BrowserFinalizingAction[] {
-	const actions: BrowserFinalizingAction[] = [];
-	if (CLICK_LIKE_COMMANDS.has(command[0] ?? "")) {
-		actions.push({
-			command,
-			kind: "click-like",
-			stepIndex,
-			targetLabel: getClickLikeTargetLabel(command, refSnapshot),
-		});
-		return actions;
-	}
-	if (command[0] === "find") {
-		const actionIndex = command.findIndex((token, index) => index >= 3 && FIND_CLICK_ACTIONS.has(token));
-		if (actionIndex !== -1) {
-			actions.push({
-				command,
-				kind: "click-like",
-				stepIndex,
-				targetLabel: getFindClickTargetLabel(command),
-			});
-		}
-		return actions;
-	}
-	const submitKey = getKeyboardSubmitKey(command)?.trim().toLowerCase();
-	if (submitKey && KEYBOARD_SUBMIT_KEYS.has(submitKey)) {
-		actions.push({
-			command,
-			kind: "keyboard-submit",
-			stepIndex,
-			targetLabel: submitKey,
-		});
-	}
-	return actions;
-}
-
-export function collectBrowserFinalizingActions(options: {
-	commandTokens: string[];
-	refSnapshot?: SessionRefSnapshot;
-	stdin?: string;
-}): BrowserFinalizingAction[] {
-	const actions = collectActionsFromCommand(options.commandTokens, options.refSnapshot);
-	if (options.commandTokens[0] !== "batch") return actions;
-	for (const { index, step } of parseValidBatchStepEntries(options.stdin)) {
-		actions.push(...collectActionsFromCommand(step, options.refSnapshot, index));
-	}
-	return actions;
-}
-
-export function shouldBlockFinalizingAction(action: BrowserFinalizingAction): boolean {
-	if (action.kind === "keyboard-submit") return true;
-	return matchesFinalActionLabel(action.targetLabel);
-}
-
-export function findBlockedFinalizingAction(options: {
-	commandTokens: string[];
-	refSnapshot?: SessionRefSnapshot;
-	stdin?: string;
-}): BrowserFinalizingAction | undefined {
-	for (const action of collectBrowserFinalizingActions(options)) {
-		if (!shouldBlockFinalizingAction(action)) continue;
-		return action;
-	}
-	return undefined;
-}