test: drive the EIP-1193 dApp approval round trips in the browser (closes #183)
All checks were successful
check / check (push) Successful in 29s

The dApp signing path was the largest unverified surface in the milestone: the
only place where the content script, the inpage provider, the background worker
and the popup all have to work together, with unit tests covering each side in
isolation and none covering the seam.

A page served by the harness speaks EIP-1193 to the real provider -- asserted by
EIP-6963 object identity, not by shape -- and eth_requestAccounts, personal_sign,
eth_signTypedData_v4 and eth_sendTransaction are each driven through to approval
and to rejection.

Every signature is recovered and compared to the approved address; the broadcast
transaction is parsed from the bytes captured at eth_sendRawTransaction and
checked for signer, recipient, value, calldata and chain. A signature that
merely came back would pass against a wrong key, a wrong message or a wrong
chain, so each assertion was demonstrated failing against a variant that is
wrong in exactly one of those ways.

The password is asserted absent from every message crossing the extension
boundary, which gives #157's fix a permanent floor rather than a one-time
review.

Two defects this surfaced are tracked separately: EIP-1193 error codes never
reach the page (#274), and approving a site connection races the popup teardown
(#275). Neither is asserted as correct here. A real dApp with real funds against
mainnet remains an uncovered human pass and is documented as such.
This commit was merged in pull request #273.
This commit is contained in:
2026-08-12 13:23:57 +02:00
parent e4c3708b84
commit d5595c0151
4 changed files with 1150 additions and 1 deletions

View File

@@ -169,6 +169,34 @@ reserve while sitting on the same side of the estimate, so swapping the two in
what [#154](https://git.eeqj.de/sneak/AutistMask/issues/154) was, and it was
previously correct by reading only.
It also covers the **dApp approval round trips** — the one place where the
content script, the inpage provider, the background worker and the approval
popup all have to work together. A local test page is served by the route
handler on a reserved-TLD origin, gets `window.ethereum` from the shipped
`MAIN`-world content script like any other page, and drives
`eth_requestAccounts`, `personal_sign`, `eth_signTypedData_v4` and
`eth_sendTransaction` through the real prompts. Every signature is recovered in
the runner and compared against the active address, the transaction assertions
run against the raw signed transaction captured at `eth_sendRawTransaction`
rather than against anything the extension reported, rejecting each prompt is
required to return a rejection to the page rather than hang or resolve, and the
password is required to be absent from every message the approval window sends
to the background — with the message that would carry it required to be present,
so that check cannot pass by observing nothing. That last one is the standing
floor under [#157](https://git.eeqj.de/sneak/AutistMask/issues/157).
Three limits of that coverage, none of them papered over. The RPC is stubbed
throughout, so this is **not** a real dApp against a real network with real
funds; that remains a human pass before 1.0.0. The site-connection prompt is
raised through `chrome.action.openPopup()`, and headless Chromium's
browser-action popup is not a page Playwright can see or click, so that one
prompt is driven at the URL the extension itself puts on the action — the same
page and the same approval id, but whether a real toolbar click shows it is not
observable here. And the EIP-1193 error code does not survive the last hop: the
rejection that crosses the boundary carries code 4001 and is asserted to, but
`src/content/inpage.js` rebuilds it as `new Error(message)`, so the calling page
catches an error with no `code` property.
Any test that drives a failure path on purpose declares the `console.error` it
is about to provoke, via `errors.expect()`. That is not a mute: the declaration
consumes exactly one matching record, and a declaration nothing matched fails

11
TODO.md
View File

@@ -45,6 +45,17 @@ undefined identifiers, which is how
# Completed Steps
- 2026-08-12: The dApp approval round trips are driven end to end in the
browser. A test page served by the harness speaks EIP-1193 to the real inpage
provider through the real content script, background worker and approval popup
for `eth_requestAccounts`, `personal_sign`, `eth_signTypedData_v4` and
`eth_sendTransaction`. Every signature is recovered and compared against the
active address, the transaction is checked against the bytes handed to the
stubbed RPC, each rejection must reach the page as a rejection, and the
password must appear in no message the approval window sends — the assertion
that gives [#157](https://git.eeqj.de/sneak/AutistMask/issues/157) a permanent
floor. This does not discharge a real dApp with real funds against mainnet
([#183](https://git.eeqj.de/sneak/AutistMask/issues/183)).
- 2026-08-12: The known-symbol spoof rule now judges the symbol a user actually
sees. `isSpoofedSymbol()` normalizes before the lookup — NFKC, then every
character that paints nothing removed (the format and default-ignorable

View File

@@ -23,6 +23,8 @@
"use strict";
const { Transaction } = require("ethers");
// Fictional ERC-20 used to seed the transaction-detail test. The symbol
// must not collide with any entry in src/shared/tokenList.js, or
// isSpoofedSymbol() in src/shared/transactions.js drops the transfer as a
@@ -62,6 +64,84 @@ function word(value) {
return "0x" + BigInt(value).toString(16).padStart(64, "0");
}
// -------------------------------------------------------- dApp fixture
//
// The origin the EIP-1193 test page is served from, and the page itself.
//
// It is a fixture like every other one in this file: the route handler
// fulfils the navigation from the string below, so the page never comes
// from a remote origin and nothing about the dApp round trips leaves the
// container. `.test` is reserved by RFC 6761 and has no owner to reach in
// the first place; the launch arguments map every host to NOTFOUND anyway.
//
// What the page deliberately does NOT do is load a provider. window.ethereum
// is put there by the shipped manifest's MAIN-world content script, exactly
// as it is on any http(s) page a user visits, so what these tests speak to
// is the real inpage provider and not a copy the harness wired up.
const DAPP_ORIGIN = "https://dapp.e2e.test";
const DAPP_URL = DAPP_ORIGIN + "/";
// Requests are parked rather than awaited. An approval prompt only exists
// while its call is in flight, so a test that awaited the promise could
// never drive the popup that has to settle it; start() files the promise
// under a key and settle() collects it once the prompt has been dealt with.
//
// The rejection branch records `code` as it arrives. EIP-1193 says a user
// rejection is a ProviderRpcError carrying code 4001; what the page can
// actually see is recorded here rather than assumed, and asserted in run.js.
//
// The message log is the page's half of the boundary observation: every
// AUTISTMASK_* message that crosses between this page and the content
// script, in both directions, verbatim.
const DAPP_HTML = [
"<!doctype html>",
'<html lang="en">',
"<head>",
'<meta charset="utf-8">',
"<title>AutistMask e2e dApp</title>",
// Inline and empty: without it Chromium asks for /favicon.ico, which
// the unstubbed-request guard would report as escaping traffic.
'<link rel="icon" href="data:,">',
"</head>",
"<body>",
"<h1>AutistMask e2e dApp</h1>",
"<script>",
"window.__dapp = {",
" messages: [],",
" calls: {},",
" start: function (key, method, params) {",
" window.__dapp.calls[key] = window.ethereum",
" .request({ method: method, params: params })",
" .then(",
" function (result) {",
" return { settled: 'resolved', result: result };",
" },",
" function (error) {",
" return {",
" settled: 'rejected',",
" message: String((error && error.message) || error),",
" hasCode: !!error && 'code' in Object(error),",
" code: error ? error.code : undefined,",
" };",
" },",
" );",
" },",
" settle: function (key) {",
" return window.__dapp.calls[key];",
" },",
"};",
"window.addEventListener('message', function (event) {",
" if (event.source !== window) return;",
" var d = event.data;",
" if (!d || typeof d.type !== 'string') return;",
" if (d.type.indexOf('AUTISTMASK') !== 0) return;",
" window.__dapp.messages.push(d);",
"});",
"</script>",
"</body>",
"</html>",
].join("\n");
// ------------------------------------------------------------ fee fixture
//
// The confirmation screen carries two different numbers for the same
@@ -101,6 +181,11 @@ const RPC_RESULTS = {
eth_estimateGas: hex(GAS_LIMIT),
eth_getTransactionCount: "0x0",
eth_maxPriorityFeePerGas: hex(PRIORITY_FEE_WEI),
// "not mined yet", which is what a node answers for a transaction it has
// only just accepted. The wait screen the dApp transaction approval hands
// off to polls this every 10 seconds; leaving it unstubbed would report
// the poll as escaping traffic the moment a test outlived one tick.
eth_getTransactionReceipt: null,
};
// The "latest" block, which ethers' getFeeData() reads baseFeePerGas from
@@ -264,6 +349,31 @@ function rpcReply(req, opts, report) {
if (req.method === "eth_getBlockByNumber") {
return Object.assign(envelope, { result: latestBlock() });
}
// The end of the dApp transaction round trip: the raw signed transaction
// the background hands to the node. It is recorded verbatim so a test can
// recover the signer from the exact bytes that were broadcast, rather than
// from anything the extension reported about them.
//
// The reply must be the transaction's real hash. ethers compares the hash
// the node returns against the one it computes itself and throws on a
// mismatch, so a constant here would fail the broadcast for a reason that
// has nothing to do with what is being tested.
if (req.method === "eth_sendRawTransaction") {
const raw = Array.isArray(req.params) ? req.params[0] : null;
let parsed;
try {
parsed = Transaction.from(raw);
} catch {
report("eth_sendRawTransaction with an undecodable transaction");
return Object.assign(envelope, {
error: { code: -32000, message: "undecodable transaction" },
});
}
if (Array.isArray(opts.broadcastTransactions)) {
opts.broadcastTransactions.push(raw);
}
return Object.assign(envelope, { result: parsed.hash });
}
if (req.method === "eth_estimateGas" && opts.failGasEstimate) {
// A refusal the node itself would produce, not a transport error:
// this is the shape the confirmation screen has to turn into
@@ -375,6 +485,8 @@ function traceEnabled(raw) {
* node-side refusal.
* @param {boolean} [opts.holdGasEstimate] hold every batch containing an
* eth_estimateGas until this is cleared again.
* @param {string[]} [opts.broadcastTransactions] every raw signed
* transaction handed to eth_sendRawTransaction, appended in order.
* @returns {Promise<{waitForServiceWorkerTraffic: (ms: number) =>
* Promise<string|null>}>}
*/
@@ -420,6 +532,18 @@ async function installNetworkStubs(ctx, opts) {
return handleRpc(route, req.postData(), opts, report);
}
// The local EIP-1193 test page. Served from here so the dApp round
// trips run against a real http(s) origin — which is what makes the
// shipped content scripts inject at all — without any remote origin
// being involved.
if (url.origin === DAPP_ORIGIN && p === "/") {
return route.fulfill({
status: 200,
contentType: "text/html; charset=utf-8",
body: DAPP_HTML,
});
}
// Blockscout v2
if (p.includes("/api/v2/")) {
if (/\/addresses\/0x[0-9a-fA-F]{40}\/transactions$/.test(p)) {
@@ -508,6 +632,8 @@ async function installNetworkStubs(ctx, opts) {
module.exports = {
installNetworkStubs,
DAPP_ORIGIN,
DAPP_URL,
FEE_ESTIMATE_WEI,
FEE_RESERVE_WEI,
STUB_COUNTERPARTY,

File diff suppressed because it is too large Load Diff