Merge pull request 'chore: release 1.2.0' (#14) from develop into main

Reviewed-on: https://git.zavage.net/Zavage-Software/search-with-chatgpt-powered-by-openai-extension/pulls/14

.gitea/workflows/firefox-extension.yml

@@ -2,13 +2,31 @@ name: Firefox Extension CI
 
 on:
   pull_request:
+    # edited: removing the WIP: prefix starts the real run.
+    types: [opened, synchronize, reopened, edited]
   push:
     branches:
       - develop
       - main
+  workflow_dispatch:
+
+# The gate and concurrency use the github.* context, Gitea's alias for gitea.*:
+# it is the form the pattern was canaried with and the one pr-flow detects.
+concurrency:
+  group: ${{ github.workflow }}-${{ github.ref }}
+  # Cancel superseded PR runs, never a push run: main's push uploads the package.
+  cancel-in-progress: ${{ github.event_name == 'pull_request' }}
 
 jobs:
   validate:
+    # WIP gate: skip PRs titled WIP: so early PRs cost no run per push. Gate on
+    # the title and nothing else, since a skipped job reports success on the
+    # required check (see ~/.claude/docs/gitea-ci-pattern.md).
+    if: >-
+      github.event_name != 'pull_request' || (
+        !startsWith(github.event.pull_request.title, 'WIP:') &&
+        !startsWith(github.event.pull_request.title, '[WIP]')
+      )
     runs-on: ubuntu-latest
     steps:
       - name: Check out repository

CHANGELOG.md

@@ -2,28 +2,34 @@
 
 All notable changes to this project will be documented in this file. See [commit-and-tag-version](https://github.com/absolute-version/commit-and-tag-version).
 
-## Unreleased
+## [1.2.0] (2026-10-18)
+
+The first update on addons.mozilla.org since 1.0. It also delivers everything
+in 1.1.0, which was never published there.
 
 - Add an Ask ChatGPT toolbar button: use selected text or open a blank chat,
   honoring saved prompt and tab/window preferences. No new permissions.
+  Requires Firefox 112 or newer, like the other new entry points.
 - Add Off / Simple / Advanced context-menu modes with migration from the old
   visibility preference. Advanced provides selection presets and link/page URL
   actions, with optional translation and a preferred language or language dialog.
 - Preserve dedicated action instructions when shortening selections; ask before
   sending shortened text and reject URLs that cannot be sent intact.
+- Stay responsive with very large selections from the shortcut or toolbar: only
+  the part that can fit in the ChatGPT link is processed.
 - Add accessible action dialogs with cancellation and retry, remove the temporary
   model/effort URL probe, and document URL sharing. No new permissions.
 - Show up front, in the options panel and README, that ChatGPT may require
   signing in and that an unsent prompt needs Enter or Send.
-
+- Limit the prompt template to 1,000 characters, so an oversized template can
-
+  no longer stop every other setting from saving.
- * Options instructions now show the configured keyboard shortcut, update when
+- Show the configured keyboard shortcut in the options instructions, updating
-   it changes, and indicate when no shortcut is assigned. Added a button to
+  when it changes and indicating when none is assigned, with a button to open
-   open Firefox’s shortcut editor where supported.
+  Firefox’s shortcut editor where supported.
- * Shortened the context menu label to “Ask ChatGPT”.
+- Shorten the context menu label to “Ask ChatGPT”.
- * Added quick usage instructions and guidance for changing the search keyword
+- Add quick usage instructions and guidance for changing the search keyword
   and default engine in Firefox to the options panel.
- * Added a “Feedback or report a bug” link to the options panel and README.
+- Add a “Feedback or report a bug” link to the options panel and README.
 
 ## [1.1.0] (2026-07-30)
 
@@ -44,7 +50,7 @@ Adds an options panel and two new ways to reach ChatGPT.
  * Permissions added: storage, menus, activeTab. None of them adds a line to
    Firefox's install prompt.
 
-## [1.0.0]) (2024-06-09)
+## [1.0.0] (2024-06-09)
 
 First release of software.
 

docs/superpowers/specs/2026-07-29-extension-options-design.md

@@ -96,6 +96,21 @@ send. The hypothesis that address-bar navigation might still be favoured
 (`Sec-Fetch-Site: none` being the most-trusted value) **does not hold in
 practice**.
 
+**Correction, 2026-10-18: `q` auto-submits again.**
+- **Signed out:** in headless Firefox 157, `chatgpt.com/?q=` sent the prompt
+  and ChatGPT answered with no keypress. This happened even when an unrelated
+  page navigated to the URL (`Sec-Fetch-Site: cross-site`).
+- **Maintainer's Firefox:** the same day, the maintainer reported that prompts
+  were sent without pressing Enter.
+- **A week earlier:** on 2026-10-12 the signed-out path showed a verification
+  error and the signed-in path still needed Enter
+  (`docs/toolbar-verification.md`).
+
+So this changed within a week and may differ between accounts. Treat both
+behaviours as possible. The copy rule below still applies, and the README
+already says ChatGPT decides whether a prompt is sent. Appendix B covers what
+this means for selection smuggling.
+
 No copy in the options panel, README, or AMO listing may say the query is
 *searched*, *sent*, or *asked* — only that ChatGPT opens **with the query filled
 in**. The two-keypress behaviour is an upstream ChatGPT change, not a defect in
@@ -698,7 +713,7 @@ by running commands against copies under `/tmp`; the repo itself was unmodified.
 | Parameter | Rating | Basis |
 |---|---|---|
 | `q` prefills | **Solid** | Dated 2026 sources; maintained tooling; OpenAI's own extension uses it; a live probe returned HTTP 200 with all parameters preserved. |
-| `q` auto-submits | **Refuted, and confirmed by first-hand test** | Prefill-only in every dated report since 2025-11-11. Tenable TRA-2025-22 documents OpenAI gating auto-submit on `Sec-Fetch-Site`. Reproduced 2026-07-29 in the maintainer's logged-in Firefox via this extension's own search-provider path. |
+| `q` auto-submits | **Varies; auto-submitting again as of 2026-10-18** | Prefill-only in every dated report from 2025-11-11, when Tenable TRA-2025-22 documented OpenAI gating auto-submit on `Sec-Fetch-Site`, and reproduced 2026-07-29 in the maintainer's logged-in Firefox. On 2026-10-18 it auto-submitted signed out, including after a cross-site navigation, and the maintainer reported the same in their own Firefox. See the correction under the `q` section. |
 | `model` honoured alongside `q` | **Uncertain** | An unresolved 2024-12-28 report says `q` forces GPT-4o regardless; counter-evidence exists from 2024–2025; no 2026 source positively confirms. Maintained tooling grades it experimental. |
 | Valid `model` slugs | **Uncertain** | No authoritative list; sightings rot (`gpt-4`, `o1-preview`, `gpt-4o`, `gpt-5-thinking`, `auto`, `instant`, `thinking`). |
 | `temporary-chat=true` | **Uncertain** | Community-confirmed originally; multiple reports of it not applying or reverting after the first message; OpenAI reworking the feature. |
@@ -790,6 +805,16 @@ composer until a human presses Enter. **Auto-submit deletes that checkpoint.**
 This is also why the context menu and shortcut remain safe *as long as they only
 prefill*.
 
+> **2026-10-18:** ChatGPT itself now auto-submits `?q=` prompts, at least for
+> some accounts (see the correction under the `q` section). While that holds,
+> the composer checkpoint above is gone for the context menu, toolbar button and
+> shortcut alike: hidden text in a selection can reach an answered prompt
+> without review. The extension still adds no submission of its own, and this
+> appendix's rejection stands. 1.2.0 ships with the behaviour known. Possible
+> mitigations, such as framing the default Ask prompt as data the way the preset
+> actions already do, or an option to review page-sourced text first, are open
+> follow-ups rather than part of this design.
+
 **`webNavigation.transitionType` is not a security boundary.** The signal looks
 right — a `gpt` search emits `transitionType: "generated"` with
 `transitionQualifiers: ["from_address_bar"]`, and content cannot forge those

lib/actions.js

@@ -13,7 +13,6 @@ const TITLES = {
 };
 
 export function getContext(info) {
-  if (info.inputType === 'password') return null;
   if (typeof info.selectionText === 'string' && info.selectionText.trim()) {
     return { kind: 'selection', source: info.selectionText.trim() };
   }

lib/build-url.js

@@ -12,9 +12,6 @@ const BASE = "https://chatgpt.com/";
 
 export const MAX_URL_CHARS = 8000;
 
-// Fraction of the kept text within which a whitespace cut is preferred.
-const WHITESPACE_WINDOW = 0.9;
-
 export function applyTemplate(template, query) {
   if (query === "") {
     return "";
@@ -42,65 +39,9 @@ function addOptionalParams(url, settings) {
   return url;
 }
 
-function lastWhitespaceIndex(text) {
+// Encode a complete prompt without silently shortening it; prepareAction owns
-  for (let i = text.length - 1; i >= 0; i -= 1) {
+// the length budget and asks before shortening. q goes first, matching the
-    if (/\s/.test(text[i])) {
+// most-cited working examples.
-      return i;
-    }
-  }
-  return -1;
-}
-
-// Returns the longest prefix of `prompt` whose finished URL fits the budget.
-// Slicing happens by CODE POINT on the decoded string, so a surrogate pair can
-// never be split and a percent-escape can never be severed (encoding happens
-// afterwards, via URLSearchParams).
-function fitToBudget(prompt, settings) {
-  const codePoints = Array.from(prompt);
-
-  const urlLengthFor = (count) => {
-    const url = new URL(BASE);
-    url.searchParams.set("q", codePoints.slice(0, count).join(""));
-    addOptionalParams(url, settings);
-    return url.href.length;
-  };
-
-  if (urlLengthFor(codePoints.length) <= MAX_URL_CHARS) {
-    return prompt;
-  }
-
-  let low = 0;
-  let high = codePoints.length;
-  while (low < high) {
-    const mid = Math.floor((low + high + 1) / 2);
-    if (urlLengthFor(mid) <= MAX_URL_CHARS) {
-      low = mid;
-    } else {
-      high = mid - 1;
-    }
-  }
-
-  const kept = codePoints.slice(0, low).join("");
-  const boundary = Math.floor(kept.length * WHITESPACE_WINDOW);
-  const cut = lastWhitespaceIndex(kept);
-  return cut >= boundary && cut > 0 ? kept.slice(0, cut) : kept;
-}
-
-export function buildChatGptUrl(settings, rawQuery) {
-  const query = String(rawQuery ?? "").trim();
-  const prompt = fitToBudget(applyTemplate(settings.promptTemplate, query), settings);
-
-  // q first, matching the most-cited working examples. Total serialized length
-  // is order-independent, so measuring without q above is sound.
-  const url = new URL(BASE);
-  if (prompt !== "") {
-    url.searchParams.set("q", prompt);
-  }
-  addOptionalParams(url, settings);
-  return url.href;
-}
-
-// Encode a complete prompt without silently shortening it.
 export function serializePrompt(settings, prompt) {
   const url = new URL(BASE);
   if (prompt) url.searchParams.set("q", prompt);

lib/prepare-action.js

@@ -25,12 +25,23 @@ export function prepareAction(request) {
   }
   const resolved = { ...request, language };
   const serialize = text => serializePrompt(request.settings, renderPrompt(resolved, text));
+  // Each code point adds at least one URL character, so none past the first
+  // MAX_URL_CHARS can fit. Capping first keeps a select-all on a huge page from
+  // costing seconds of re-encoding. A capped source is never sent as ready: its
+  // kept prefix may trim down to something that fits and silently drop the rest.
+  const points = [];
+  let capped = false;
+  for (const point of source) {
+    if (points.length === MAX_URL_CHARS) { capped = true; break; }
+    points.push(point);
+  }
+  if (!capped) {
     const url = serialize(source);
     if (url.length <= MAX_URL_CHARS) return { kind: 'ready', url };
+  }
   if (isUrl) return error('This URL is too long to send intact. No part of it has been sent.');
   const marker = '\n[Selected text shortened]';
   if (serialize(marker).length >= MAX_URL_CHARS) return error('The prompt instructions are too long. Shorten your prompt template and try again.');
-  const points = Array.from(source);
   let low = 0, high = points.length;
   while (low < high) {
     const mid = Math.ceil((low + high) / 2);

manifest.json

@@ -1,10 +1,10 @@
 {
   "name": "Search with ChatGPT powered by OpenAI!",
-  "description": "Adds a ChatGPT search engine",
+  "description": "Adds a ChatGPT search engine, plus a right-click menu, toolbar button and shortcut that open ChatGPT with your selected text.",
   "manifest_version": 2,
   "author": "Mathew Guest <mat@zavage.net>",
   "homepage_url": "https://zavage-software.com/portfolio/search-with-ChatGPT",
-  "version": "1.1",
+  "version": "1.2",
   "icons": {
     "16": "icons/powered-by-openai-icon-16x16.png",
     "32": "icons/powered-by-openai-icon-32x32.png",

package-lock.json

@@ -1,12 +1,12 @@
 {
   "name": "search-with-chatgpt-powered-by-openai-extension",
-  "version": "1.0.0",
+  "version": "1.2.0",
   "lockfileVersion": 3,
   "requires": true,
   "packages": {
     "": {
       "name": "search-with-chatgpt-powered-by-openai-extension",
-      "version": "1.0.0",
+      "version": "1.2.0",
       "devDependencies": {
         "web-ext": "10.5.0"
       },

package.json

@@ -1,6 +1,6 @@
 {
   "name": "search-with-chatgpt-powered-by-openai-extension",
-  "version": "1.1.0",
+  "version": "1.2.0",
   "private": true,
   "type": "module",
   "description": "Development tooling for the Search with ChatGPT Firefox extension.",

test/actions.test.js

@@ -9,7 +9,6 @@ test('context prioritizes selection over link and excludes unrelated targets', (
   assert.deepEqual(actions.getContext({selectionText:' ',linkUrl:'https://example.com'}), {kind:'link',source:'https://example.com'});
   assert.equal(actions.getContext({mediaType:'image',pageUrl:'https://example.com'}), null);
   assert.equal(actions.getContext({editable:true,pageUrl:'https://example.com'}), null);
-  assert.equal(actions.getContext({selectionText:'secret',inputType:'password'}), null);
   assert.equal(actions.getContext({pageUrl:'https://example.com'}).kind, 'page');
 });
 test('selection actions and optional translation reflect settings', () => {

test/build-url.test.js

@@ -2,168 +2,86 @@ import test from "node:test";
 import assert from "node:assert/strict";
 
 import { DEFAULT_SETTINGS } from "../lib/defaults.js";
-import { buildChatGptUrl, MAX_URL_CHARS } from "../lib/build-url.js";
+import { applyTemplate, serializePrompt } from "../lib/build-url.js";
 
 const withSettings = (overrides) => ({ ...DEFAULT_SETTINGS, ...overrides });
 const queryOf = (url) => new URL(url).searchParams.get("q");
 const paramsOf = (url) => [...new URL(url).searchParams.keys()];
 
 test("defaults produce only the q parameter on the root path", () => {
-  assert.equal(
+  assert.equal(serializePrompt(DEFAULT_SETTINGS, "hello"), "https://chatgpt.com/?q=hello");
-    buildChatGptUrl(DEFAULT_SETTINGS, "hello"),
-    "https://chatgpt.com/?q=hello",
-  );
 });
 
-test("an empty or whitespace-only query omits q entirely", () => {
+test("an empty prompt omits q entirely", () => {
-  for (const raw of ["", "   ", "\n\t", null, undefined]) {
+  assert.equal(serializePrompt(DEFAULT_SETTINGS, ""), "https://chatgpt.com/");
-    const url = buildChatGptUrl(DEFAULT_SETTINGS, raw);
+  for (const template of ["{query}", "Explain: {query}", "Be concise:", ""]) {
-    assert.equal(paramsOf(url).includes("q"), false, `raw=${JSON.stringify(raw)}`);
+    assert.equal(applyTemplate(template, ""), "", `template=${JSON.stringify(template)}`);
-    assert.equal(url, "https://chatgpt.com/");
   }
 });
 
 test("legacy model values never add a model URL parameter", () => {
-  assert.equal(buildChatGptUrl(withSettings({ model: "gpt-5.6-sol" }), "hi"),
+  assert.equal(serializePrompt(withSettings({ model: "gpt-5.6-sol" }), "hi"),
-    buildChatGptUrl(DEFAULT_SETTINGS, "hi"));
+    serializePrompt(DEFAULT_SETTINGS, "hi"));
 });
 
 test("temporaryChat and webSearch each appear only when enabled", () => {
-  assert.equal(paramsOf(buildChatGptUrl(DEFAULT_SETTINGS, "hi")).includes("temporary-chat"), false);
+  assert.deepEqual(paramsOf(serializePrompt(DEFAULT_SETTINGS, "hi")), ["q"]);
-  assert.equal(paramsOf(buildChatGptUrl(DEFAULT_SETTINGS, "hi")).includes("hints"), false);
 
-  const temp = new URL(buildChatGptUrl(withSettings({ temporaryChat: true }), "hi"));
+  const temp = new URL(serializePrompt(withSettings({ temporaryChat: true }), "hi"));
   assert.equal(temp.searchParams.get("temporary-chat"), "true");
 
-  const search = new URL(buildChatGptUrl(withSettings({ webSearch: true }), "hi"));
+  const search = new URL(serializePrompt(withSettings({ webSearch: true }), "hi"));
   assert.equal(search.searchParams.get("hints"), "search");
 });
 
-test("all three URL-affecting settings combine, q first", () => {
+test("all URL-affecting settings combine, q first", () => {
-  const url = buildChatGptUrl(
+  const settings = withSettings({ temporaryChat: true, webSearch: true, promptTemplate: "Explain: {query}" });
-    withSettings({
+  const url = serializePrompt(settings, applyTemplate(settings.promptTemplate, "entropy"));
-      temporaryChat: true,
-      webSearch: true,
-      promptTemplate: "Explain: {query}",
-    }),
-    "entropy",
-  );
   assert.deepEqual(paramsOf(url), ["q", "temporary-chat", "hints"]);
   assert.equal(queryOf(url), "Explain: entropy");
 });
 
 test("openIn never influences the URL", () => {
-  const base = buildChatGptUrl(DEFAULT_SETTINGS, "hi");
+  const base = serializePrompt(DEFAULT_SETTINGS, "hi");
   for (const openIn of ["new-tab", "background-tab", "new-window"]) {
-    assert.equal(buildChatGptUrl(withSettings({ openIn }), "hi"), base);
+    assert.equal(serializePrompt(withSettings({ openIn }), "hi"), base);
   }
 });
 
 test("special characters survive encoding round-trip", () => {
   const raw = 'a & b # c ? d = e + f "g" <h> 100% 😀 日本語\nnewline';
-  const url = buildChatGptUrl(DEFAULT_SETTINGS, raw);
+  const url = serializePrompt(DEFAULT_SETTINGS, raw);
   assert.equal(queryOf(url), raw);
   assert.equal(url.includes(" "), false);
   assert.equal(url.includes("#"), false);
 });
 
 test("the template replaces every {query} occurrence", () => {
-  const url = buildChatGptUrl(withSettings({ promptTemplate: "{query} / {query}" }), "x");
+  assert.equal(applyTemplate("{query} / {query}", "x"), "x / x");
-  assert.equal(queryOf(url), "x / x");
 });
 
 test("a template without {query} is treated as a prefix, never dropping the query", () => {
-  const url = buildChatGptUrl(withSettings({ promptTemplate: "Be concise:" }), "why");
+  assert.equal(applyTemplate("Be concise:", "why"), "Be concise: why");
-  assert.equal(queryOf(url), "Be concise: why");
 });
 
 test("an empty or whitespace-only template yields the bare query", () => {
-  for (const promptTemplate of ["", "   "]) {
+  for (const template of ["", "   "]) {
-    assert.equal(queryOf(buildChatGptUrl(withSettings({ promptTemplate }), "why")), "why");
+    assert.equal(applyTemplate(template, "why"), "why");
   }
 });
 
 test("a non-string template degrades to the bare query", () => {
-  assert.equal(queryOf(buildChatGptUrl(withSettings({ promptTemplate: null }), "why")), "why");
+  assert.equal(applyTemplate(null, "why"), "why");
-});
-
-test("an over-long prompt is truncated to fit the URL budget", () => {
-  const url = buildChatGptUrl(DEFAULT_SETTINGS, "x".repeat(20000));
-  assert.ok(url.length <= MAX_URL_CHARS, `length was ${url.length}`);
-  assert.ok(queryOf(url).length > 0);
-  assert.ok(queryOf(url).length < 20000);
-});
-
-test("truncation accounts for the other parameters", () => {
-  const settings = withSettings({ temporaryChat: true, webSearch: true });
-  const url = buildChatGptUrl(settings, "x".repeat(20000));
-  assert.ok(url.length <= MAX_URL_CHARS, `length was ${url.length}`);
-  assert.equal(new URL(url).searchParams.get("hints"), "search");
-});
-
-test("truncation never splits a surrogate pair", () => {
-  const url = buildChatGptUrl(DEFAULT_SETTINGS, "😀".repeat(9000));
-  assert.ok(url.length <= MAX_URL_CHARS);
-  const lonelySurrogate = /[\uD800-\uDBFF](?![\uDC00-\uDFFF])|(?<![\uD800-\uDBFF])[\uDC00-\uDFFF]/;
-  assert.equal(lonelySurrogate.test(queryOf(url)), false);
-});
-
-test("truncation never leaves a severed percent-escape", () => {
-  const url = buildChatGptUrl(DEFAULT_SETTINGS, "😀 ".repeat(9000));
-  assert.equal(/%[0-9A-Fa-f]?$/.test(url), false);
-  assert.doesNotThrow(() => decodeURIComponent(new URL(url).search.slice(3)));
-});
-
-test("truncation prefers a whitespace boundary, so no word is cut in half", () => {
-  const url = buildChatGptUrl(DEFAULT_SETTINGS, "lorem ipsum ".repeat(2000));
-  const tokens = queryOf(url).trim().split(/\s+/);
-  assert.ok(tokens.length > 1);
-  for (const token of tokens) {
-    assert.ok(token === "lorem" || token === "ipsum", `truncated mid-word: ${token}`);
-  }
-});
-
-test("a query needing no truncation is left byte-identical", () => {
-  const raw = "a modest question about thermodynamics";
-  assert.equal(queryOf(buildChatGptUrl(DEFAULT_SETTINGS, raw)), raw);
 });
 
 // String.prototype.replaceAll gives $&, $`, $', and $$ special meaning in the
 // REPLACEMENT string, and `query` (arbitrary page-selected text) is exactly
 // that argument. These selections are ordinary shell one-liners, not
 // adversarial input, and must round-trip byte-identically into `q`.
-test("a selection containing $& survives the default {query} template byte-identically", () => {
+for (const raw of ["awk print $& here", "shell $` and $' here", "PID is $$ in bash"]) {
-  const raw = "awk print $& here";
+  test(`a selection containing ${raw.match(/\$./g).join(" and ")} survives templating byte-identically`, () => {
-  const url = buildChatGptUrl(DEFAULT_SETTINGS, raw);
+    assert.equal(applyTemplate("{query}", raw), raw);
-  assert.equal(queryOf(url), raw);
+    assert.equal(applyTemplate("Explain: {query}", raw), `Explain: ${raw}`);
-});
+    assert.equal(queryOf(serializePrompt(DEFAULT_SETTINGS, applyTemplate("{query}", raw))), raw);
-
-test("a selection containing $& survives a prefix template byte-identically", () => {
-  const raw = "awk print $& here";
-  const url = buildChatGptUrl(withSettings({ promptTemplate: "Explain: {query}" }), raw);
-  assert.equal(queryOf(url), `Explain: ${raw}`);
-});
-
-test("a selection containing $` and $' survives the default {query} template byte-identically", () => {
-  const raw = "shell $` and $' here";
-  const url = buildChatGptUrl(DEFAULT_SETTINGS, raw);
-  assert.equal(queryOf(url), raw);
-});
-
-test("a selection containing $` and $' survives a prefix template byte-identically", () => {
-  const raw = "shell $` and $' here";
-  const url = buildChatGptUrl(withSettings({ promptTemplate: "Explain: {query}" }), raw);
-  assert.equal(queryOf(url), `Explain: ${raw}`);
-});
-
-test("a selection containing $$ survives the default {query} template byte-identically", () => {
-  const raw = "PID is $$ in bash";
-  const url = buildChatGptUrl(DEFAULT_SETTINGS, raw);
-  assert.equal(queryOf(url), raw);
-});
-
-test("a selection containing $$ survives a prefix template byte-identically", () => {
-  const raw = "PID is $$ in bash";
-  const url = buildChatGptUrl(withSettings({ promptTemplate: "Explain: {query}" }), raw);
-  assert.equal(queryOf(url), `Explain: ${raw}`);
   });
+}

test/prepare-action.test.js

@@ -1,6 +1,7 @@
 import test from 'node:test';
 import assert from 'node:assert/strict';
 import { DEFAULT_SETTINGS } from '../lib/defaults.js';
+import { MAX_URL_CHARS } from '../lib/build-url.js';
 const { prepareAction } = await import('../lib/prepare-action.js').catch(()=>({}));
 const req = (overrides={}) => ({actionId:'ask',source:'Hello',settings:{...DEFAULT_SETTINGS},...overrides});
 test('ready actions preserve source and hints, translation asks only when needed', () => {
@@ -23,6 +24,17 @@ test('URLs are validated and preserved whole, never shortened', () => {
   assert.equal(out.kind,'ready');
   assert.ok(new URL(out.url).searchParams.get('q').endsWith(source));
 });
+test('shortened selections end on a whole word and decode cleanly', () => {
+  assert.equal(typeof prepareAction, 'function');
+  const out=prepareAction(req({source:'lorem ipsum '.repeat(2000)}));
+  assert.equal(out.kind,'shortened');
+  const words=new URL(out.url).searchParams.get('q').replace('[Selected text shortened]','').trim().split(/\s+/);
+  assert.ok(words.length>1);
+  for (const word of words) assert.ok(word==='lorem'||word==='ipsum',`cut mid-word: ${word}`);
+  const emoji=prepareAction(req({source:'😀 '.repeat(9000)}));
+  assert.equal(emoji.kind,'shortened');
+  assert.doesNotThrow(()=>decodeURIComponent(new URL(emoji.url).search.slice(1)));
+});
 test('long selections preserve instructions, repetition and Unicode with explicit consent', () => {
   assert.equal(typeof prepareAction, 'function');
   const settings={...DEFAULT_SETTINGS,promptTemplate:'START {query} AGAIN {query} END',temporaryChat:true,webSearch:true};
@@ -38,3 +50,22 @@ test('long selections preserve instructions, repetition and Unicode with explici
   const translated=prepareAction(req({actionId:'translate',language:'Spanish',source:'x'.repeat(20000)}));
   assert.equal(translated.kind,'shortened');assert.match(new URL(translated.url).searchParams.get('q'), /Spanish/);
 });
+test('huge selections are capped before shortening, with identical output', () => {
+  assert.equal(typeof prepareAction, 'function');
+  // Select-all on a very large page; shortening this took over a second uncapped.
+  const source='😀 word '.repeat(625000);
+  const started=performance.now();
+  const out=prepareAction(req({source}));
+  const elapsed=performance.now()-started;
+  assert.equal(out.kind,'shortened');
+  assert.ok(elapsed<250,`took ${elapsed.toFixed(0)} ms`);
+  const capped=Array.from(source).slice(0,MAX_URL_CHARS).join('');
+  assert.deepEqual(prepareAction(req({source:capped})),out);
+});
+test('a selection too long to send whole always needs consent, even if its start would fit', () => {
+  assert.equal(typeof prepareAction, 'function');
+  // The first MAX_URL_CHARS code points trim down to "a"; sending just that would drop the rest unasked.
+  const out=prepareAction(req({source:'a'+' '.repeat(MAX_URL_CHARS)+'b'}));
+  assert.equal(out.kind,'shortened');
+  assert.ok(new URL(out.url).searchParams.get('q').includes('[Selected text shortened]'));
+});