pi-browser-harness 0.6.0 → 0.8.0

package/src/domains/form.ts

@@ -0,0 +1,270 @@
+import { Type } from "typebox";
+import type { BrowserClient } from "../client";
+import { safeJs } from "../util/js-template";
+import { type Result, err, ok } from "../util/result";
+import { defineBrowserTool, type ToolErr, type ToolOk } from "../util/tool";
+import { interactiveDiff, resolveRefToObjectId } from "./ref-resolve";
+
+const FillArgs = Type.Object({
+  ref: Type.Optional(
+    Type.String({
+      description:
+        "Stable element ref from browser_snapshot (e.g. 'e7'). PREFERRED over selector — survives re-renders. When set, selector is ignored.",
+    }),
+  ),
+  selector: Type.Optional(Type.String({ description: "CSS selector of the form field to fill (fallback when no ref)" })),
+  value: Type.String({ description: "Value to set in the field" }),
+});
+
+/**
+ * Run an element-scoped function either against a CSS selector (querySelector +
+ * apply) or a ref (resolve to objectId, callFunctionOn with element as `this`).
+ * The shared `fnBody` is a function body where `this` is the target element and
+ * the single argument is the value — identical logic on both paths. Returns the
+ * function's return value (parsed) or a ToolErr.
+ *
+ * On the ref path a detached node yields a stale-ref error; on the selector path
+ * a missing element returns { status: "not_found" } from the function body.
+ */
+const runOnElement = async (
+  client: BrowserClient,
+  opts: { ref?: string | undefined; selector?: string | undefined; fnBody: string; arg: unknown },
+): Promise<Result<unknown, ToolErr>> => {
+  if (opts.ref !== undefined) {
+    const objectId = await resolveRefToObjectId(client, opts.ref);
+    if (!objectId.success) return objectId;
+    const r = await client.session().call("Runtime.callFunctionOn", {
+      objectId: objectId.data,
+      functionDeclaration: `function (arg) { ${opts.fnBody} }`,
+      arguments: [{ value: opts.arg }],
+      returnByValue: true,
+    });
+    if (!r.success) return err({ kind: "cdp_error", message: r.error.message });
+    return ok((r.data as { result?: { value?: unknown } }).result?.value);
+  }
+  if (opts.selector === undefined) {
+    return err({ kind: "invalid_state", message: "Provide either `ref` or `selector`." });
+  }
+  // Selector path: bind the same trusted function body to the matched element
+  // via .call(). fnBody is harness-authored (not user input); only the selector
+  // and value are interpolated, and those go through safeJs.
+  const prelude = safeJs`
+    (() => {
+      const el = document.querySelector(${opts.selector});
+      if (!el) return { status: "not_found" };
+      const __arg = ${opts.arg};
+      return (function (arg) {`;
+  const expr = `${prelude} ${opts.fnBody} }).call(el, __arg); })()`;
+  const r = await client.evaluateJs(expr);
+  if (!r.success) return err({ kind: "cdp_error", message: r.error.message });
+  return ok(r.data);
+};
+
+/**
+ * Result shape returned by the in-page fill script. Discriminated by `status`
+ * so the handler can map page-level outcomes to typed ToolErr kinds.
+ */
+type FillResult =
+  | { status: "ok"; tag: string; kind: string; value: string }
+  | { status: "not_found" }
+  | { status: "not_fillable"; tag: string };
+
+// Element-scoped fill logic. `this` is the target element, `arg` is the value.
+// Write via the native prototype value setter so React's _valueTracker
+// registers the change — assigning el.value directly does NOT, which is why
+// controlled inputs revert on re-render. Shared verbatim by the ref path
+// (callFunctionOn) and the selector path (querySelector + .call).
+const FILL_FN = `
+  const el = this;
+  const tag = el.tagName;
+  const proto =
+    el instanceof HTMLTextAreaElement ? HTMLTextAreaElement.prototype
+    : el instanceof HTMLInputElement ? HTMLInputElement.prototype
+    : null;
+  if (proto) {
+    const desc = Object.getOwnPropertyDescriptor(proto, "value");
+    const setter = desc && desc.set;
+    if (setter) setter.call(el, arg);
+    else el.value = arg;
+    el.dispatchEvent(new Event("input", { bubbles: true }));
+    el.dispatchEvent(new Event("change", { bubbles: true }));
+    return { status: "ok", tag, kind: el.type || tag.toLowerCase(), value: el.value };
+  }
+  if (el.isContentEditable) {
+    // Rich-text editors (Slack/Notion/ProseMirror) listen for beforeinput/input.
+    el.focus();
+    let done = false;
+    try {
+      const sel = window.getSelection();
+      if (sel) { sel.removeAllRanges(); const rng = document.createRange(); rng.selectNodeContents(el); sel.addRange(rng); }
+      done = document.execCommand("insertText", false, arg);
+    } catch (e) { done = false; }
+    if (!done) {
+      el.textContent = arg;
+      el.dispatchEvent(new InputEvent("input", { bubbles: true }));
+    }
+    return { status: "ok", tag, kind: "contenteditable", value: el.textContent || "" };
+  }
+  return { status: "not_fillable", tag };
+`;
+
+export const fillTool = defineBrowserTool({
+  name: "browser_fill",
+  label: "Browser Fill",
+  description:
+    "Fill a form field (input/textarea/contenteditable). PREFERRED: pass `ref` (e.g. 'e7') from browser_snapshot — survives re-renders. Fallback: a CSS `selector`. Writes through the native value setter and fires bubbling 'input'/'change' events, so React/Vue/Angular controlled components and rich-text editors update correctly — unlike browser_type. Returns the field's value after writing, plus a compact diff of page changes.",
+  promptSnippet: "Fill a form field by ref (preferred) or selector (works with React/Vue controlled inputs)",
+  promptGuidelines: [
+    "PREFER `ref` from browser_snapshot (the '[eN]' handle) over a guessed CSS selector — refs survive re-renders and don't require you to invent a selector.",
+    "Pass the desired value; no browser_click is needed first.",
+    "A 'ref is stale' error means the page changed — re-run browser_snapshot to get fresh refs.",
+    "Use browser_type instead only for keystroke-sensitive widgets (autocomplete, masked/segmented inputs).",
+    "The result reports the field's value after writing and an appended page-changes diff — confirm both match what you intended before moving on.",
+  ],
+  parameters: FillArgs,
+  async handler(args, { client }): Promise<Result<ToolOk, ToolErr>> {
+    const r = await runOnElement(client, { ref: args.ref, selector: args.selector, fnBody: FILL_FN, arg: args.value });
+    if (!r.success) return r;
+    const res = r.data as FillResult | undefined;
+    const target = args.ref ?? args.selector ?? "";
+    if (res === undefined || res.status === "not_found") {
+      return err({
+        kind: "invalid_state",
+        message: `No element matched: ${target}`,
+        details: { ref: args.ref, selector: args.selector },
+      });
+    }
+    if (res.status === "not_fillable") {
+      return err({
+        kind: "invalid_state",
+        message: `Element <${res.tag.toLowerCase()}> is not fillable (not an input, textarea, or contenteditable): ${target}. For <select> use browser_select_option.`,
+        details: { ref: args.ref, selector: args.selector, tag: res.tag },
+      });
+    }
+    const diff = await interactiveDiff(client);
+    return ok({
+      text: `Filled ${target} = ${JSON.stringify(res.value)}${diff}`,
+      details: { ref: args.ref, selector: args.selector, value: args.value, verified: res.value, tag: res.tag, kind: res.kind },
+    });
+  },
+});
+
+const FocusArgs = Type.Object({
+  ref: Type.Optional(
+    Type.String({ description: "Stable element ref from browser_snapshot (e.g. 'e7'). PREFERRED over selector." }),
+  ),
+  selector: Type.Optional(Type.String({ description: "CSS selector of the element to focus (fallback when no ref)" })),
+});
+
+type FocusResult = { status: "ok"; tag: string } | { status: "not_found" } | { status: "not_focusable"; tag: string };
+
+const FOCUS_FN = `
+  const el = this;
+  if (typeof el.focus !== "function") return { status: "not_focusable", tag: el.tagName };
+  el.focus();
+  return document.activeElement === el ? { status: "ok", tag: el.tagName } : { status: "not_focusable", tag: el.tagName };
+`;
+
+export const focusTool = defineBrowserTool({
+  name: "browser_focus",
+  label: "Browser Focus",
+  description:
+    "Focus an element via the DOM .focus() method — deterministic, no coordinate accuracy needed. PREFERRED: pass `ref` from browser_snapshot; fallback: a CSS `selector`. Use before browser_type when a click might miss the field.",
+  promptSnippet: "Focus an element by ref (preferred) or selector",
+  promptGuidelines: [
+    "PREFER `ref` from browser_snapshot over a CSS selector — survives re-renders.",
+    "Use before browser_type to guarantee the right field is focused without relying on click coordinates.",
+    "For simply setting a value, prefer browser_fill (it doesn't need a separate focus step).",
+  ],
+  parameters: FocusArgs,
+  async handler(args, { client }): Promise<Result<ToolOk, ToolErr>> {
+    const r = await runOnElement(client, { ref: args.ref, selector: args.selector, fnBody: FOCUS_FN, arg: "" });
+    if (!r.success) return r;
+    const res = r.data as FocusResult | undefined;
+    const target = args.ref ?? args.selector ?? "";
+    if (res === undefined || res.status === "not_found") {
+      return err({ kind: "invalid_state", message: `No element matched: ${target}`, details: { ref: args.ref, selector: args.selector } });
+    }
+    if (res.status === "not_focusable") {
+      return err({ kind: "invalid_state", message: `Element <${res.tag.toLowerCase()}> could not be focused: ${target}`, details: { ref: args.ref, selector: args.selector, tag: res.tag } });
+    }
+    return ok({ text: `Focused ${target}`, details: { ref: args.ref, selector: args.selector, tag: res.tag } });
+  },
+});
+
+const SelectOptionArgs = Type.Object({
+  ref: Type.Optional(
+    Type.String({ description: "Stable element ref of the <select> from browser_snapshot. PREFERRED over selector." }),
+  ),
+  selector: Type.Optional(Type.String({ description: "CSS selector of the <select> element (fallback when no ref)" })),
+  value: Type.Optional(Type.String({ description: "Option value attribute to select" })),
+  label: Type.Optional(Type.String({ description: "Visible option text to select (exact match)" })),
+  index: Type.Optional(Type.Integer({ minimum: 0, description: "Zero-based option index to select" })),
+});
+
+type SelectResult =
+  | { status: "ok"; value: string; label: string }
+  | { status: "not_found" }
+  | { status: "not_select"; tag: string }
+  | { status: "no_match"; options: ReadonlyArray<{ value: string; label: string }> };
+
+// `this` is the <select>; `arg` is { wv, wl, wi } (null = not provided).
+const SELECT_FN = `
+  const el = this;
+  if (el.tagName !== "SELECT") return { status: "not_select", tag: el.tagName };
+  const opts = Array.from(el.options);
+  const wv = arg.wv, wl = arg.wl, wi = arg.wi;
+  let opt = null;
+  if (wi !== null) opt = opts[wi] || null;
+  else if (wv !== null) opt = opts.find(o => o.value === wv) || null;
+  else if (wl !== null) opt = opts.find(o => o.text === wl) || null;
+  if (!opt) return { status: "no_match", options: opts.map(o => ({ value: o.value, label: o.text })) };
+  el.value = opt.value;
+  el.dispatchEvent(new Event("input", { bubbles: true }));
+  el.dispatchEvent(new Event("change", { bubbles: true }));
+  return { status: "ok", value: opt.value, label: opt.text };
+`;
+
+export const selectOptionTool = defineBrowserTool({
+  name: "browser_select_option",
+  label: "Browser Select Option",
+  description:
+    "Select an option in a native <select> element by value, visible label, or index. PREFERRED: pass `ref` from browser_snapshot; fallback: a CSS `selector`. Sets the selection and fires bubbling 'input'/'change' events so framework listeners update. On no match, returns the available options.",
+  promptSnippet: "Select an option in a native <select> by ref (preferred) or selector",
+  promptGuidelines: [
+    "PREFER `ref` from browser_snapshot over a CSS selector — survives re-renders.",
+    "Provide exactly one of value, label, or index to choose the option.",
+    "For custom (non-native) dropdowns built from divs, click to open then browser_click the option instead.",
+    "If no option matches, the error details list the available options so you can retry.",
+  ],
+  parameters: SelectOptionArgs,
+  async handler(args, { client }): Promise<Result<ToolOk, ToolErr>> {
+    if (args.value === undefined && args.label === undefined && args.index === undefined) {
+      return err({ kind: "invalid_state", message: "Provide one of value, label, or index." });
+    }
+    // null sentinels distinguish "not provided" from a real empty-string match.
+    const arg = { wv: args.value ?? null, wl: args.label ?? null, wi: args.index ?? null };
+    const r = await runOnElement(client, { ref: args.ref, selector: args.selector, fnBody: SELECT_FN, arg });
+    if (!r.success) return r;
+    const res = r.data as SelectResult | undefined;
+    const target = args.ref ?? args.selector ?? "";
+    if (res === undefined || res.status === "not_found") {
+      return err({ kind: "invalid_state", message: `No element matched: ${target}`, details: { ref: args.ref, selector: args.selector } });
+    }
+    if (res.status === "not_select") {
+      return err({ kind: "invalid_state", message: `Element <${res.tag.toLowerCase()}> is not a <select>: ${target}`, details: { ref: args.ref, selector: args.selector, tag: res.tag } });
+    }
+    if (res.status === "no_match") {
+      return err({
+        kind: "invalid_state",
+        message: `No matching option in ${target}. Available: ${res.options.map((o) => o.label).join(", ")}`,
+        details: { ref: args.ref, selector: args.selector, options: res.options },
+      });
+    }
+    const diff = await interactiveDiff(client);
+    return ok({
+      text: `Selected "${res.label}" (value=${JSON.stringify(res.value)}) in ${target}${diff}`,
+      details: { ref: args.ref, selector: args.selector, value: res.value, label: res.label },
+    });
+  },
+});

package/src/domains/keyboard.ts

@@ -3,6 +3,7 @@ import { safeJs } from "../util/js-template";
 import { virtualKeyCode } from "../util/keycodes";
 import { type Result, err, ok } from "../util/result";
 import { defineBrowserTool, type ToolErr, type ToolOk } from "../util/tool";
+import { resolveRefToObjectId } from "./ref-resolve";
 
 const TypeArgs = Type.Object({
   text: Type.String({ description: "Text to type" }),
@@ -19,6 +20,29 @@ export const typeTool = defineBrowserTool({
   ],
   parameters: TypeArgs,
   async handler(args, { client }): Promise<Result<ToolOk, ToolErr>> {
+    // Guard against silent text loss: Input.insertText succeeds regardless of
+    // whether an editable element is focused, so without this check typed text
+    // can vanish into the void with no error. Fail loudly with an actionable
+    // message instead. IIFE starts with "(" so evaluateJs returns it directly.
+    const focused = await client.evaluateJs(safeJs`
+      (() => {
+        const el = document.activeElement;
+        const ok = !!el && el !== document.body &&
+          (el.tagName === "INPUT" || el.tagName === "TEXTAREA" || el.isContentEditable);
+        return { ok, tag: el ? el.tagName : null };
+      })()
+    `);
+    if (focused.success) {
+      const f = focused.data as { ok: boolean; tag: string | null };
+      if (!f.ok) {
+        return err({
+          kind: "invalid_state",
+          message:
+            "No editable element is focused — typed text would be lost. Click the field with browser_click first, or use browser_fill with a selector.",
+          details: { activeElement: f.tag },
+        });
+      }
+    }
     const r = await client.session().call("Input.insertText", { text: args.text });
     if (!r.success) return err({ kind: "cdp_error", message: r.error.message });
     return ok({ text: `Typed: "${args.text}"` });
@@ -67,6 +91,14 @@ export const pressKeyTool = defineBrowserTool({
     };
     const down = await client.session().call("Input.dispatchKeyEvent", downParams);
     if (!down.success) return err({ kind: "cdp_error", message: down.error.message });
+    // For a printable character with no command modifier (Ctrl/Meta/Alt = 1|2|4),
+    // emit a `char` event so the page receives keypress/textInput and the
+    // character is actually inserted. Shift (8) is allowed (capitals/symbols).
+    const hasCommandModifier = (modifiers & (1 | 2 | 4)) !== 0;
+    if (isChar && !hasCommandModifier) {
+      const charEv = await client.session().call("Input.dispatchKeyEvent", { type: "char", text: k, key: k });
+      if (!charEv.success) return err({ kind: "cdp_error", message: charEv.error.message });
+    }
     const up = await client.session().call("Input.dispatchKeyEvent", { type: "keyUp", key: k, code: k, modifiers });
     if (!up.success) return err({ kind: "cdp_error", message: up.error.message });
     return ok({ text: `Pressed: ${k}${modifiers ? ` (modifiers=${modifiers})` : ""}` });
@@ -74,7 +106,10 @@ export const pressKeyTool = defineBrowserTool({
 });
 
 const DispatchKeyArgs = Type.Object({
-  selector: Type.String({ description: "CSS selector of the target element" }),
+  ref: Type.Optional(
+    Type.String({ description: "Stable element ref from browser_snapshot (e.g. 'e7'). PREFERRED over selector." }),
+  ),
+  selector: Type.Optional(Type.String({ description: "CSS selector of the target element (fallback when no ref)" })),
   key: Type.String({ description: "Key value (e.g., 'Enter', 'a')" }),
   eventType: Type.Optional(
     Type.Union(
@@ -88,17 +123,38 @@ export const dispatchKeyTool = defineBrowserTool({
   name: "browser_dispatch_key",
   label: "Browser Dispatch Key",
   description:
-    "Dispatch a DOM KeyboardEvent on a specific element via JS injection. Use for React/Vue components that listen to synthetic events more reliably than CDP input.",
-  promptSnippet: "Dispatch a DOM KeyboardEvent on a specific element",
+    "Dispatch a DOM KeyboardEvent on a specific element via JS injection. PREFERRED: pass `ref` from browser_snapshot; fallback: a CSS `selector`. Use for React/Vue components that listen to synthetic events more reliably than CDP input.",
+  promptSnippet: "Dispatch a DOM KeyboardEvent on an element by ref (preferred) or selector",
   promptGuidelines: [
-    "Dispatches a synthetic DOM KeyboardEvent — for React/Vue synthetic event listeners. Does NOT insert text into inputs (use browser_type or browser_press_key for actual typing).",
+    "PREFER `ref` from browser_snapshot over a CSS selector — survives re-renders.",
+    "Dispatches a synthetic DOM KeyboardEvent — for React/Vue synthetic event listeners. Does NOT insert text (use browser_type or browser_press_key for actual typing).",
     "Try browser_press_key first; only use browser_dispatch_key when the page ignores raw CDP key events.",
-    "The selector must match exactly one or more elements; zero matches is reported as an error.",
     "eventType defaults to 'keydown'.",
   ],
   parameters: DispatchKeyArgs,
   async handler(args, { client }): Promise<Result<ToolOk, ToolErr>> {
     const eventType = args.eventType ?? "keydown";
+    const target = args.ref ?? args.selector ?? "";
+    // Ref path: dispatch on the single resolved node. Selector path: dispatch on
+    // all matches (preserves the original multi-match behavior).
+    if (args.ref !== undefined) {
+      const objectId = await resolveRefToObjectId(client, args.ref);
+      if (!objectId.success) return objectId;
+      const r = await client.session().call("Runtime.callFunctionOn", {
+        objectId: objectId.data,
+        functionDeclaration: `function (type, key) { this.dispatchEvent(new KeyboardEvent(type, { key, bubbles: true, cancelable: true })); return 1; }`,
+        arguments: [{ value: eventType }, { value: args.key }],
+        returnByValue: true,
+      });
+      if (!r.success) return err({ kind: "cdp_error", message: r.error.message });
+      return ok({
+        text: `Dispatched ${eventType}('${args.key}') on ${target}`,
+        details: { matched: 1, ref: args.ref },
+      });
+    }
+    if (args.selector === undefined) {
+      return err({ kind: "invalid_state", message: "Provide either `ref` or `selector`." });
+    }
     const expr = safeJs`
       (() => {
         const els = document.querySelectorAll(${args.selector});

package/src/domains/page.ts

@@ -86,3 +86,48 @@ export const waitForLoadTool = defineBrowserTool({
     return err({ kind: "timeout", message: `Page did not finish loading in ${args.timeout ?? 15}s` });
   },
 });
+
+const WaitForArgs = Type.Object({
+  selector: Type.Optional(Type.String({ description: "CSS selector to wait for (or wait to disappear with gone:true)" })),
+  text: Type.Optional(Type.String({ description: "Wait until this text appears anywhere in document.body" })),
+  gone: Type.Optional(Type.Boolean({ default: false, description: "If true with a selector, wait until it is absent instead of present" })),
+  timeout: Type.Optional(Type.Number({ default: 5, minimum: 1, maximum: 60, description: "Max seconds to wait. Default: 5." })),
+});
+
+export const waitForTool = defineBrowserTool({
+  name: "browser_wait_for",
+  label: "Browser Wait For",
+  description:
+    "Poll until a condition holds: a selector appears (or disappears with gone:true), or text appears in the page. Returns a typed timeout error if the deadline elapses. Use to wait for SPA content to render before interacting.",
+  promptSnippet: "Wait for an element or text to appear (or an element to disappear)",
+  promptGuidelines: [
+    "Use before browser_fill / browser_click on dynamic pages so you act after the element renders.",
+    "Provide selector (default: wait until present; gone:true: wait until absent) or text.",
+    "Returns a typed timeout error if the condition isn't met within the timeout.",
+  ],
+  parameters: WaitForArgs,
+  async handler(args, { client, signal }): Promise<Result<ToolOk, ToolErr>> {
+    if (args.selector === undefined && args.text === undefined) {
+      return err({ kind: "invalid_state", message: "Provide a selector or text to wait for." });
+    }
+    const gone = args.gone ?? false;
+    const timeoutMs = (args.timeout ?? 5) * 1000;
+    const expr =
+      args.selector !== undefined
+        ? safeJs`(() => { const found = !!document.querySelector(${args.selector}); return ${gone} ? !found : found; })()`
+        : safeJs`(() => { const t = document.body ? document.body.innerText : ""; return t.includes(${args.text}); })()`;
+    const start = Date.now();
+    while (Date.now() - start < timeoutMs) {
+      if (signal?.aborted) return err({ kind: "internal", message: "aborted" });
+      const r = await client.evaluateJs(expr);
+      if (r.success && r.data === true) {
+        const ms = Date.now() - start;
+        const what = args.selector !== undefined ? `${args.selector}${gone ? " gone" : ""}` : `text "${args.text}"`;
+        return ok({ text: `Condition met (${what}) in ${Math.round(ms / 100) / 10}s`, details: { ms } });
+      }
+      await sleep(50, signal);
+    }
+    const what = args.selector !== undefined ? `${args.selector}${gone ? " to disappear" : ""}` : `text "${args.text}"`;
+    return err({ kind: "timeout", message: `Timed out after ${args.timeout ?? 5}s waiting for ${what}` });
+  },
+});

package/src/domains/ref-resolve.ts

@@ -0,0 +1,152 @@
+import type { BrowserClient } from "../client";
+import { type Result, err, ok } from "../util/result";
+import type { ToolErr } from "../util/tool";
+import { buildTree, collectInteractiveTargets, type RawAxNode, type SlimNode } from "./snapshot";
+
+type Box = { x: number; y: number; width: number; height: number; cx: number; cy: number };
+
+const staleErr = (ref: string): ToolErr => ({
+  kind: "invalid_state",
+  message: `Ref ${ref} is unknown or stale — re-run browser_snapshot to get fresh refs.`,
+  details: { ref },
+});
+
+/**
+ * Resolve a ref to its CDP objectId (a handle to the live JS node). Use when a
+ * tool needs to run JS against the element (fill, focus, select, dispatch_key).
+ * Re-resolves through the backendNodeId at call time, so it reflects the current
+ * DOM even after re-renders — and fails loudly if the node was detached.
+ */
+export const resolveRefToObjectId = async (
+  client: BrowserClient,
+  ref: string,
+): Promise<Result<string, ToolErr>> => {
+  const backendId = client.session().resolveRef(ref);
+  if (backendId === undefined) return err(staleErr(ref));
+  const resolved = await client.session().call("DOM.resolveNode", { backendNodeId: backendId });
+  if (!resolved.success) {
+    // A detached node (removed by a re-render) fails to resolve — surface it as
+    // a stale ref rather than a raw CDP error so the agent knows to re-snapshot.
+    return err(staleErr(ref));
+  }
+  const objectId = (resolved.data as { object?: { objectId?: string } }).object?.objectId;
+  if (!objectId) return err(staleErr(ref));
+  return ok(objectId);
+};
+
+/** Resolve a ref to its current backendNodeId (for CDP calls that take one, e.g. DOM.setFileInputFiles). */
+export const resolveRefToBackendId = (client: BrowserClient, ref: string): Result<number, ToolErr> => {
+  const backendId = client.session().resolveRef(ref);
+  if (backendId === undefined) return err(staleErr(ref));
+  return ok(backendId);
+};
+
+/**
+ * Resolve a ref to the current center of its bounding box, in viewport CSS
+ * pixels. Use for browser_click. Fetched fresh via DOM.getBoxModel, so a
+ * re-rendered/moved element is clicked at its new position.
+ */
+export const resolveRefToBox = async (
+  client: BrowserClient,
+  ref: string,
+): Promise<Result<Box, ToolErr>> => {
+  const backendId = client.session().resolveRef(ref);
+  if (backendId === undefined) return err(staleErr(ref));
+  const r = await client.session().call("DOM.getBoxModel", { backendNodeId: backendId });
+  if (!r.success) return err(staleErr(ref));
+  const data = r.data as { model?: { content?: ReadonlyArray<number>; width?: number; height?: number } };
+  const content = data.model?.content;
+  if (!content || content.length < 8) return err(staleErr(ref));
+  const x = content[0]!;
+  const y = content[1]!;
+  const width = data.model!.width ?? 0;
+  const height = data.model!.height ?? 0;
+  if (width <= 0 || height <= 0) return err(staleErr(ref));
+  return ok({
+    x: Math.round(x),
+    y: Math.round(y),
+    width: Math.round(width),
+    height: Math.round(height),
+    cx: Math.round(x + width / 2),
+    cy: Math.round(y + height / 2),
+  });
+};
+
+const sigOf = (n: SlimNode): string => `${n.role}|${n.name ?? ""}|${n.value ?? ""}|${n.state ?? ""}`;
+const keyOf = (n: SlimNode): string => `${n.role}|${n.name ?? ""}`;
+
+/**
+ * Recompute the interactive elements after a mutation and return a compact,
+ * human/LLM-readable diff vs. the prior snapshot's baseline — changed values,
+ * newly-appeared elements (marked `*`), and removed ones. Re-publishes the fresh
+ * ref map so subsequent ref-based calls target the post-mutation DOM.
+ *
+ * Deliberately cheap: it re-reads the AX tree (no per-node box fetch — the
+ * expensive part of a full snapshot) and emits a few lines, not a full tree, so
+ * a long action sequence doesn't blow up the context.
+ *
+ * Returns "" on any failure or when nothing changed — callers append it only if
+ * non-empty, so a diff is best-effort and never blocks the primary action.
+ */
+export const interactiveDiff = async (client: BrowserClient): Promise<string> => {
+  const session = client.session();
+  const prev = session.refSignatures();
+
+  const axRes = await session.call("Accessibility.getFullAXTree", {});
+  if (!axRes.success) return "";
+  const rawNodes = (axRes.data as { nodes: ReadonlyArray<RawAxNode> }).nodes;
+  const slim = buildTree(rawNodes, { interestingOnly: true, maxNodes: 1000 });
+  const targets = collectInteractiveTargets(slim);
+
+  // Re-publish fresh refs so later ref calls resolve against the new DOM.
+  const refMap = new Map<string, number>();
+  const refSig = new Map<string, string>();
+  targets.forEach(({ node, backendId }, i) => {
+    const ref = `e${i + 1}`;
+    node.ref = ref;
+    refMap.set(ref, backendId);
+    refSig.set(ref, sigOf(node));
+  });
+  session.setRefMap(refMap, refSig);
+
+  // Compare by stable identity (role|name) rather than ref position, since refs
+  // renumber when the element set changes. Build prev/current identity → value.
+  const prevByKey = new Map<string, string>();
+  for (const sig of prev.values()) {
+    const parts = sig.split("|");
+    prevByKey.set(`${parts[0]}|${parts[1]}`, sig);
+  }
+  const curByKey = new Map<string, { ref: string; sig: string }>();
+  for (const { node } of targets) {
+    if (node.ref) curByKey.set(keyOf(node), { ref: node.ref, sig: sigOf(node) });
+  }
+
+  const appeared: string[] = [];
+  const changed: string[] = [];
+  for (const [key, { ref, sig }] of curByKey) {
+    const before = prevByKey.get(key);
+    if (before === undefined) {
+      appeared.push(`  *[${ref}] ${key.replace("|", ' "')}"`);
+    } else if (before !== sig) {
+      const va = before.split("|")[2] ?? "";
+      const vb = sig.split("|")[2] ?? "";
+      const sa = before.split("|")[3] ?? "";
+      const sb = sig.split("|")[3] ?? "";
+      const detail = va !== vb ? `value ${JSON.stringify(va)} → ${JSON.stringify(vb)}` : `state ${JSON.stringify(sa)} → ${JSON.stringify(sb)}`;
+      changed.push(`  [${ref}] ${key.replace("|", ' "')}": ${detail}`);
+    }
+  }
+  const removed: string[] = [];
+  for (const key of prevByKey.keys()) {
+    if (!curByKey.has(key)) removed.push(`  ${key.replace("|", ' "')}"`);
+  }
+
+  if (appeared.length === 0 && changed.length === 0 && removed.length === 0) return "";
+  const lines: string[] = ["", "Page changes (re-snapshot for full tree):"];
+  if (changed.length) lines.push("Changed:", ...changed.slice(0, 12));
+  if (appeared.length) lines.push("New (*):", ...appeared.slice(0, 12));
+  if (removed.length) lines.push("Removed:", ...removed.slice(0, 8));
+  const overflow = Math.max(0, changed.length - 12) + Math.max(0, appeared.length - 12) + Math.max(0, removed.length - 8);
+  if (overflow > 0) lines.push(`  …and ${overflow} more (browser_snapshot for full state)`);
+  return lines.join("\n");
+};