PLI: return the 7-byte location object, bound every proactive chain

The PLI answer used the 5-byte form 13 05 <MCC/MNC> <LAC> with LAC 0000 and no Cell ID. That is the *Network Rejection event download* form (TS 31.111 8.19.1); a PLI location for GERAN is 7 bytes - MCC/MNC, LAC and Cell ID. A card applet that re-reads its location (the Alfa MultiIMSI applet, after a *LOCI status change) treated it as empty and re-requested PLI on every STATUS tick. The default is now 13 07 00F110 0001 0001: PLMN 001/01 (the test network), LAC 0001, Cell ID 0001. _build_tr injects it after the Result; the TR data is pinned by a test. While verifying a live PLI, the manual /api/status-poll chain turned out to be unbounded: it looped on the PLI storm while holding _CARD_LOCK, so the request never returned and every card operation waited. _handle_proactive_chain's max_rounds now defaults to _POLL_CHAIN_MAX (the background poll always passed it; the manual poll, ENVELOPE, menu and TP chains did not), and the manual poll passes it explicitly. A card that re-queues a command can no longer wedge the card lock from any call site. Tests: the PLI default + the built TR, the default chain bound and the manual-poll call pin. 1061 Python + 1069 frontend; sw cache simple-v477.
test runner: add the stuck-card watchdog
2026-10-24 01:51:35 +03:00 · 2026-10-24 00:43:39 +03:00 · 2026-10-23 23:44:42 +03:00 · 2026-10-23 21:46:10 +03:00 · 2026-10-23 21:44:45 +03:00 · 2026-10-23 21:41:30 +03:00
22 changed files with 2004 additions and 116 deletions
@@ -9,6 +9,90 @@ every change carries a descriptive commit message.

 ## [Unreleased]

+## [3.37.0] — 2026-10-23
+
+### Added
+
+- **Load testing** for test scripts and suites.  `POST /api/test/run` takes an optional
+  `iterations` count (the PWA's **Iterations** field next to *Start* / *Run suite*): the
+  run repeats N times, successful iterations are discarded (only the iteration counter and
+  the timing are kept; the per-step log is not rendered or retained), and the first
+  iteration that ends in an **error** stops the loop and keeps that iteration's full
+  report.  A manual stop ends it after the current iteration, and a card session change
+  between iterations is an error (a swapped card is never run against another card's
+  script).  The SCP80 counter state is shared across iterations, so packets continue from
+  what the previous run consumed.  The run status gains a `load` block (`iteration`,
+  `done`, `failed_iteration`, `elapsed_ms`, `avg_ms`); the count is a run parameter, not
+  part of the stored script or suite.
+
+- The **stuck-card watchdog** for test runs.  A card that stops answering (no status
+  word, no driver error) used to leave the runner inside a blocking `SCardTransmit`
+  holding the card lock — *Stop* could not take effect, the presence monitor waited for
+  the lock and a load test hung in that iteration without explanation.  After 30 s
+  without runner progress (a bounded STATUS/drain loop beats once per attempt) the
+  watchdog marks the run `stuck`, records `the card stopped responding (step N, no
+  answer for N s)`, sets the stop flag and **tries** to cancel the blocked transmit by
+  disconnecting the PC/SC link on a background thread.  Where PC/SC can cancel it, the
+  step fails through the ordinary `on_fail` path and the run finalises as an **error**
+  (never `stopped`); where it cannot (pcsc-lite serializes the disconnect behind the
+  pending transmit), the status already names the step until the driver's own card
+  timeout returns.  The transport is marked stale once the disconnect returns, so the
+  next *Equip* rebuilds it.  The status gains `stuck`; the help EN/RU,
+  `docs/test-scripts.md` and `docs/api.md` describe the behaviour and the
+  background-disconnect caveat.
+
+### Fixed
+
+- The **PLI location answer** now uses the 7-byte PLI form (TS 31.111 8.19.1):
+  PLMN 001/01, LAC `0001`, Cell ID `0001`.  The former default was the 5-byte
+  *Network Rejection event* form (no Cell ID) with LAC `0000`; a card applet
+  that re-reads its location after a `*LOCI` status change (the Alfa MultiIMSI
+  applet) treated it as empty and re-requested PLI on every STATUS tick — the
+  PLI storm that slowed RAM Explore.  The TR data is now
+  `13 07 00F110 0001 0001`.
+- The **proactive chain is bounded on every call site**: `_handle_proactive_chain`'s
+  `max_rounds` defaults to `_POLL_CHAIN_MAX` (the background poll always passed it;
+  the manual STATUS poll and the ENVELOPE/menu/TP chains did not).  A card that
+  re-queues a command (the PLI storm) could keep the manual poll looping forever
+  while it held the card lock — `/api/status-poll` never returned and every card
+  request waited.  The manual poll also passes the bound explicitly.
+- The load-test loop closes properly on a raised runner error (it used to leave the run
+  `running` forever - the PWA polled a dead run and *Clear report* stayed refused), on a
+  card session change inside a suite iteration, and on a member that was stopped; a
+  warning in any iteration is reflected in the final status, `avg_ms` covers the attempted
+  iterations (the failing one included), and each suite iteration's `counters_before` is
+  read fresh from the preset store.
+
+## [3.36.0] — 2026-10-23
+
+### Added
+
+- The batch **"waiting for the next card"** state is now a deliberate cue: the status
+  modal shows a card-into-reader animation next to the prompt, a short chime sounds
+  **once per card** when the run enters the waiting phase (Web Audio — no audio asset;
+  edge-triggered in the 2 s poll, so it does not repeat), and while the tab is hidden the
+  title hints "Insert next card".  The speaker button next to the status mutes the chime
+  and the choice is remembered (`simple_ram_batch_sound`); the chime is unlocked by the
+  Start click (browser autoplay policy); reduced-motion users get the static card-above-
+  the-slot picture.  The card follows the app's SIM silhouette (`sim.svg`): a cut corner
+  at the top and the contact pad near the bottom edge, sliding behind the reader bar.
+
+### Changed
+
+- A **batch delete that finds the object already gone** is still a success — the goal is met
+  and the run continues — but it is no longer reported as a plain "OK": the attempt's verdict
+  reads **Already absent**, the delete run's counts separate real deletions from no-ops
+  (`Deleted: N · Already absent: M`), and the notice says the card was already absent.  The
+  counts gain an `already_absent` bucket (the attempt result is `already_absent`); a
+  reinstall's tolerated pre-delete is *not* affected — its outcome is still the install.
+- The READMEs link the project's news channel: [Telegram — @youeyeseesee](https://t.me/youeyeseesee).
+
+### Fixed
+
+- The prebuilt `style.css` had gone stale: a rebuild restored utility classes the UI
+  already used but the output lacked (e.g. `bg-blue-50` on the batch notice, `inline-flex`,
+  `mx-2`, `max-h-48`, `w-44`, `min-w-24`).
+
 ## [3.35.0] — 2026-10-23

 ### Added
@@ -2,6 +2,8 @@

 > Formerly known as OTAMan — renamed to SIMple on 2026-09-19.

+📢 Project news and updates: [Telegram channel](https://t.me/youeyeseesee)
+
 SIMple is a workbench for UICC/SIM card specialists: an offline HTML/JS PWA plus a local card server ([`pysim-simple-server`](pysim_simple_server/), a pySim/PC-SC bridge that also serves the PWA). It covers the full lab cycle:

 - **APDU workbench** — SIM/USIM RFM and Expanded Script builders (TS 102 226), GlobalPlatform RAM commands, a C-APDU/R-APDU parser and a response decoder (ISO 7816 / TS 102 221 / GP);
@@ -2,6 +2,8 @@

 > Прежнее название — OTAMan; переименован в SIMple 2026-09-19.

+📢 Новости и обновления проекта: [Telegram-канал](https://t.me/youeyeseesee)
+
 SIMple — инструментарий для специалистов по UICC/SIM-картам: автономный HTML/JS PWA и локальный сервер карт ([`pysim-simple-server`](pysim_simple_server/) — мост к pySim/PC-SC, он же раздаёт PWA). Закрывает весь цикл лабораторной работы:

 - **APDU-верстак** — конструкторы SIM/USIM RFM и Expanded Script (TS 102 226), команды GlobalPlatform RAM, парсер C-APDU/R-APDU и декодер ответов (ISO 7816 / TS 102 221 / GP);
@@ -1006,11 +1006,22 @@ on every stop** as long as the card is in the reader; a card reset (session
 change) stops the suite, skips the remaining members and is flagged in the
 report.

+`iterations` (optional, 1-1,000,000, default 1) repeats the script or the
+whole suite: successful iterations are discarded (only the counter and the
+timing are kept; the per-step log is not retained), the first iteration that
+ends in an **error** stops the loop and keeps that iteration's report, a
+manual stop ends it after the current iteration, and a card session change
+between iterations is an error.  The status gains `load`:
+`{enabled, iterations, iteration, done, failed_iteration, started, finished,
+elapsed_ms, avg_ms}`.
+
 The suite run's snapshot adds `suite`: the member records (`status`, `steps`,
 `note`, `log` - the semantic run-log excerpt), the `summary` (ok/warning/
 error/skipped counts, `wall_ms`, `stopped`, `session_changed`, the SCP80
 `counters_before`/`counters_after` per keyset), the `start_reset` flag and
-the `reset` result (ICCID + ms) and the run's `adm` state.
+the `reset` result (ICCID + ms) and the run's `adm` state.  A successful load
+test leaves `suite` `null` (no iteration is tracked); a failing one keeps the
+failing iteration's suite state.
 The report is exported from the PWA as JSON or Markdown; the run state lives
 in the server until `POST /api/test/clear` (no history).

@@ -1028,7 +1039,15 @@ The current (or last) run:
 ```

 `status` becomes `ok`/`warning`/`error`/`stopped` when the run finishes;
-expected/actual pairs are reported per check.
+expected/actual pairs are reported per check.  A card that stops answering is
+not waited on silently: after 30 s without runner progress the watchdog sets
+`stuck: true`, `stop` and `error` (`the card stopped responding (step N, no
+answer for N s)`) and tries to cancel the blocked transmit by disconnecting
+the PC/SC link on a background thread (pcsc-lite serializes the disconnect
+behind the pending transmit, so the call returns only when the driver's card
+timeout fires).  Once the step returns, the run finalises as `error` - a stuck
+card is a failure, not a user stop - and the next equip rebuilds the link
+(the transport is marked stale).

 ### `POST /api/test/stop`

@@ -1750,13 +1769,16 @@ Lists the PROVIDE LOCAL INFORMATION qualifier codes with their names.

 Returns the effective PLI data dictionary as a qualifier-code map: the
 operator's override, else the built-in default (the qualifiers whose objects
-the pinned ETSI specs define - location as the test network 001/01, the test
-IMEI/IMEISV, language, access technology, search mode, battery; v3.23.1).
+the pinned ETSI specs define - location as the test network 001/01 with test
+LAC 0001 and Cell ID 0001 (the 7-byte PLI form per TS 31.111 8.19.1; the
+5-byte form without the Cell ID belongs to the Network Rejection event
+download), the test IMEI/IMEISV, language, access technology, search mode,
+battery; v3.23.1, the PLI location form fixed in v3.37.0).
 The date/time qualifier stays empty - the host clock answers it while no
 override is set.

 ```json
-{"00": "930500F1100000", "0A": "630104", "03": ""}
+{"00": "130700F11000010001", "0A": "630104", "03": ""}
 ```

 ### `POST /api/pli-dict`
@@ -229,6 +229,42 @@ TEST-RUN step 4: MENU-SELECT ENVELOPE=80c2000009d30702020181900101 item=1 ('One'
  (`APDU-TRACE(…): <cmd> → SW: … RESP: …`) for every APDU, including the
  file-manager reads and pySim's auto-handler traffic.

+## Load testing
+
+A run may be repeated: `POST /api/test/run {..., "iterations": N}` (the
+PWA's **Iterations** field next to Start / Run suite) executes the script or
+the whole suite N times.  Successful iterations are discarded - only the
+iteration counter and the timing are kept, so memory use stays flat.  The
+first iteration that ends in an error stops the loop and keeps its full
+report (steps, failed checks, log excerpt; for a suite, the member report);
+a manual stop ends it after the current iteration, and a card session change
+between iterations is an error.  The SCP80 counter state is shared across
+iterations, so packets continue from what the previous iteration consumed.
+The iteration count is a run parameter - the stored script/suite is unchanged.
+
+## A card that stops answering
+
+pyscard's `SCardTransmit` is blocking and has no timeout: a card that stops
+answering (no status word, no driver error) would leave the runner inside the
+step while it holds the card lock - Stop could not take effect, the presence
+monitor would wait for the lock, and a load test would hang in that iteration.
+The runner's **watchdog** watches a progress heartbeat: after 30 s without one
+(a bounded STATUS/drain loop beats once per attempt, so a long poll is not
+mistaken for a stuck card) it marks the run `stuck`, sets the error
+`the card stopped responding (step N, no answer for N s)` and the stop flag,
+and **tries** to cancel the blocked transmit by disconnecting the PC/SC link
+on a background thread.  Where PC/SC can cancel it, the blocking call raises,
+the step fails through the ordinary path (`on_fail` applies) and the run
+finalises as an **error** - never as `stopped`, even though the stop flag
+ended the loop.  The disconnect runs in the background because pcsc-lite
+serializes it behind the pending transmit (both take the per-context lock), so
+on that stack the call returns only when the driver's own card timeout fires
+(libccid derives it from the card's ATR); the run is already flagged and the
+status names the step meanwhile.  The transport is marked stale once the
+disconnect returns, so the next **Equip** rebuilds the link.  A physical card
+removal is unaffected: the driver aborts the transmit and the presence monitor
+clears the card session.
+
 ## Document shape

 ```json
@@ -324,13 +324,13 @@
 			<tr><td class="py-1 px-2">Install Package (.cap file)</td><td class="py-1 px-2">Sends a <code class="font-mono text-sm">.cap</code> file to the card via the server: INSTALL[for load] &rarr; LOAD &times;N &rarr; INSTALL[for install (+make selectable)]. The load file is split into LOAD APDUs of the <strong>LOAD block size</strong> (1&ndash;240 bytes of payload, 240 by default); each secured packet is delivered as a single SMS or, when larger, as a concatenated SMS-PP download of up to 5 segments, so a large <code class="font-mono text-sm">.cap</code> simply takes several SMS. As soon as the file is selected it is validated and its <strong>memory requirements are estimated</strong> in the form: the <strong>NVRAM requirement</strong> (the code image + persistent data, the GlobalPlatform Card Spec C6+C8 style total) and the RAM estimate. The same box reports the <strong>libraries the CAP is linked against</strong> (its Import component: library name, AID and the minimum version <code class="font-mono text-sm">≥ x.y</code>, with standard names and families resolved and a Java Card SDK hint where known), the package/applet identity and the component breakdown (sizes and shares in load-file order) &mdash; grouped under the <strong>Package</strong>, <strong>Requires</strong>, <strong>Memory</strong> and <strong>Components</strong> headings. The <strong>Platform</strong> line names the minimum Java Card level the card must provide (derived from the <code class="font-mono text-sm">javacard.framework</code> import; the header <code class="font-mono text-sm">int</code> flag additionally requires int support, Java Card 2.2+) and the <strong>ETSI release</strong> the UICC API imports come from (<code class="font-mono text-sm">uicc.toolkit</code> version &rarr; REL-N, per the workspace <code class="font-mono text-sm">sim-tools</code> corpus; shown with its basis in brackets): the JCRE checks every import against its own packages when it verifies the load file (JC VM spec 4.5.2), so a card whose platform is older rejects the LOAD as soon as the <strong>Import</strong> component is complete (the step line carries the card's remote SW and names the component the block completed, e.g. <em>completes Import</em>): the failure hint and the popup then point at the <strong>CAP compatibility test</strong> (one click &mdash; the same <code class="font-mono text-sm">.cap</code> is still selected) to find the missing package and version; a smaller LOAD block only helps when the block exceeds the card's buffer. A corrupt or wrong-format file fails the analysis, and <strong>Execute</strong> stays disabled until a successful analysis. The estimate is informational only and is never used for installation: NVRAM is the code image plus the persistent data (the GlobalPlatform C6+C8 style total, Table 11-48 &mdash; card memory management, rounding and the registry entry are not included, and the static field image is counted in both parts); the suggested <strong>C6</strong> follows the load file, and the suggested quotas are decimal byte counts (the hex in parentheses is how the C7/C8 install parameters encode them: C6 = non-volatile code, C7 = volatile RAM, C8 = non-volatile data).  The <strong>Delete the package before installing (with its applications)</strong> checkbox prepends a cascade GP <code class="font-mono text-sm">DELETE</code> on the package AID (the load file with its modules and applications, P2=<code class="font-mono text-sm">0x80</code>) &mdash; a leftover package blocks INSTALL[for load]; 'not found' (a fresh card) proceeds, any other delete failure stops the chain with its reason.</td></tr>
 			<tr class="border-b border-gray-200 dark:border-slate-700"><td class="py-1 px-2">Batch install (.cap file, per card)</td><td class="py-1 px-2">The bulk workflow: one <code class="font-mono text-sm">.cap</code> and one set of install parameters, applied to card after card.  <strong>Start batch</strong> needs no card &mdash; the server owns the loop: it waits for a card, installs the moment the equip finishes, then waits for the next one.  Every card's <strong>preset is matched by ICCID</strong> and supplies the keyset (the ISD role TAR's default, v3.26.0), the keys and the counter; a card with no matching preset or no usable ISD keyset is reported and skipped.  Each attempt lands in the list with its time, ICCID, preset, verdict and duration; a failure shows the failing step and the remote status word &mdash; the batch never retries and never stops on a failure (failures are yours to manage afterwards).  A card is attempted <strong>once per run</strong>: re-equipping it reports &ldquo;already processed&rdquo; &mdash; restart the batch to retry it.  <strong>Stop</strong> finishes the current card first (an interrupted LOAD chain can leave a half-loaded package) and then ends the run; <strong>Clear</strong> drops the finished report.  The loop runs in the server, so closing the tab, switching views or reloading the page does not interrupt it; a server restart ends it.  A batch and a test run exclude each other.  The batch shares the install form, so the <strong>Delete the package before installing</strong> option applies to every card.</td></tr>
 			<tr class="border-b border-gray-200 dark:border-slate-700"><td class="py-1 px-2">Batch reinstall app (per card)</td><td class="py-1 px-2">The repair workflow: for every equipped card, delete the applet instance (tolerating 'not found') and run INSTALL [for install] again from the already loaded package &mdash; no <code class="font-mono text-sm">.cap</code> and no LOAD.  The fields are the single INSTALL [for install] parameters (Load File / Module / Instance AIDs, privileges, the toolkit/STK parameters, <em>make selectable</em>); the preset, keyset and counter come from each card's own preset like the other batch operations, and the attempts list shows both verdicts (delete and install).</td></tr>
-			<tr class="border-b border-gray-200 dark:border-slate-700"><td class="py-1 px-2">Batch delete (per card)</td><td class="py-1 px-2">Deletes one object on every equipped card: an application (instance only, P2=<code class="font-mono text-sm">0x00</code>) or a package with its applications (cascade, P2=<code class="font-mono text-sm">0x80</code>).  Fields: the AID and the target.  A 'not found' result (the object is already gone) is reported as such and counted OK; any other failure is reported with the card's status word.</td></tr>
+			<tr class="border-b border-gray-200 dark:border-slate-700"><td class="py-1 px-2">Batch delete (per card)</td><td class="py-1 px-2">Deletes one object on every equipped card: an application (instance only, P2=<code class="font-mono text-sm">0x00</code>) or a package with its applications (cascade, P2=<code class="font-mono text-sm">0x80</code>).  Fields: the AID and the target.  A 'not found' result (the object is already gone) is reported as <strong>Already absent</strong> and counted separately from real deletions &mdash; it is still a success (the goal is met) and the run continues; any other failure is reported with the card's status word.</td></tr>
 			<tr class="border-b border-gray-200 dark:border-slate-700"><td class="py-1 px-2">CAP compatibility test</td><td class="py-1 px-2">Answers one question without installing anything: <strong>can this card satisfy the CAP's import list?</strong> The server runs the format check, INSTALL[for load] and the LOAD blocks only up to the block that completes the <strong>Import</strong> component &mdash; the point where the JCRE verifies the import list &mdash; then stops: no last-block flag and no INSTALL[for install], so nothing is committed and nothing has to be deleted afterwards. The <strong>Import versions to test</strong> list is pre-filled from the CAP analysis: <code class="font-mono text-sm">AID=version</code> rewrites an import's requested version, <code class="font-mono text-sm">+AID=version</code> appends a synthetic import (a card-capability query: the LOAD passes only if the card exports that package at &ge; that version), and <strong>Set all 0.0</strong> drops every requirement. Two-step use: press Set all 0.0 &mdash; if the test passes, the import list was the blocker; then restore one line at a time to find the package and version the card lacks. The verdict states it covers the <strong>LOAD/import gate only</strong>: the link gate (whether the applet's actual API calls exist) is what a real install checks at INSTALL[for install]. The run and its result appear in a popup (like the install): the step lines, the verdict and a per-AID detail block &mdash; on a rejection it names the CAP section the failing block covers (load-file byte range and components, and whether that block completes <strong>Import</strong>), the remote status word with its meaning, the import list exactly as tested (CAP versions, probe rewrites, appended synthetic entries, unmatched lines) and what the probe state implies; on success it lists the import list that passed.</td></tr>
 			<tr class="border-b border-gray-200 dark:border-slate-700"><td class="py-1 px-2">INSTALL [for install] (loaded package)</td><td class="py-1 px-2">For a package that is already loaded on the card (its Load File AID and Module AID, privileges, install/STK parameters and an optional <em>make selectable</em>).  The applet instance is deleted first (tolerating 'not found' &mdash; a fresh card just installs; any other delete failure stops the chain with its reason), so the install parameters can be iterated without re-loading the <code class="font-mono text-sm">.cap</code>.</td></tr>
 			<tr><td class="py-1 px-2">INSTALL [for make selectable]</td><td class="py-1 px-2">One INSTALL [for make selectable] APDU (the instance AID) for an INSTALLED, non-selectable applet instance.</td></tr>
 		</tbody>
 	</table>
-	<p class="text-sm mb-3">A batch operation shows its status in a modal window: the live progress (or &ldquo;Insert next card. Waiting for a card...&rdquo;) on top and the processed cards below, with <strong>Stop</strong> and <strong>Clear</strong> inside.  It opens on Start and on a page load while a run is active; <strong>Close</strong> stays disabled until the run ends &mdash; the run itself is server-side and survives a reload.  After a finished run the panel&rsquo;s <strong>Show status</strong> button reopens the report.</p>
+	<p class="text-sm mb-3">A batch operation shows its status in a modal window: the live progress (or &ldquo;Insert next card. Waiting for a card...&rdquo;) on top and the processed cards below, with <strong>Stop</strong> and <strong>Clear</strong> inside.  It opens on Start and on a page load while a run is active; <strong>Close</strong> stays disabled until the run ends &mdash; the run itself is server-side and survives a reload.  After a finished run the panel&rsquo;s <strong>Show status</strong> button reopens the report.  While the run waits for a card the modal shows a card-into-reader animation and plays a short chime once per card (the speaker button next to the status mutes it and the choice is remembered); when the window is in the background the tab title asks for the next card.</p>

 	<h4 id="ram-explorer" class="font-medium mb-1">Explorer View</h4>
 	<p class="text-sm mb-2">After &ldquo;Explore Card&rdquo; runs, the explorer displays:</p>
@@ -539,7 +539,7 @@
 		<li><strong>06</strong> Access Technology &middot; <strong>08</strong> IMEISV &middot; <strong>09</strong> Search Mode</li>
 		<li><strong>0A</strong> Battery &middot; <strong>0E</strong> Multiple Access Technologies</li>
 	</ul>
-	<p class="text-sm mb-3">Values persist server-side until restart.  Sensible defaults are pre-settled for the qualifiers the ETSI specs define (location as the test network 001/01, the test IMEI 490154203237518 and its IMEISV, language &lsquo;en&rsquo;, E-UTRAN access technology, automatic search mode, a full battery): a value you set overrides its default, and <strong>clearing the field restores it</strong>; the date/time qualifier is answered from the host clock while empty, and the 3GPP/3GPP2-specific qualifiers (NMR, ESN, MEID, CSG/HNB, WLAN, slices, CAG, &hellip;) stay empty - the ETSI specs define no contents for them.  A field the server cannot use (a value that is not hex, an odd number of digits, a code outside the qualifier list) is named under the Save button instead of being silently dropped.  When the card issues PLI, the server injects the values into the TERMINAL RESPONSE - also into a test-script expectation's answer.</p>
+	<p class="text-sm mb-3">Values persist server-side until restart.  Sensible defaults are pre-settled for the qualifiers the ETSI specs define (location as the test network 001/01 with test LAC 0001 and Cell ID 0001 &mdash; the 7-byte PLI form, not the shorter Network Rejection one, the test IMEI 490154203237518 and its IMEISV, language &lsquo;en&rsquo;, E-UTRAN access technology, automatic search mode, a full battery): a value you set overrides its default, and <strong>clearing the field restores it</strong>; the date/time qualifier is answered from the host clock while empty, and the 3GPP/3GPP2-specific qualifiers (NMR, ESN, MEID, CSG/HNB, WLAN, slices, CAG, &hellip;) stay empty - the ETSI specs define no contents for them.  A field the server cannot use (a value that is not hex, an odd number of digits, a code outside the qualifier list) is named under the Save button instead of being silently dropped.  When the card issues PLI, the server injects the values into the TERMINAL RESPONSE - also into a test-script expectation's answer.</p>

 	<h3 id="network-simulation" class="text-lg font-medium mb-2">8.7 Network simulation</h3>
 	<p class="text-sm mb-2">Replays the write patterns a real phone performs when the network condition changes (trace study in <code class="font-mono text-sm">projects/UICC_NAA.md</code>): EPS attach, service loss / limited service, roaming denial, 2G fallback, SMS delivery, cell-broadcast reconfiguration and AUTHENTICATE. One button per scenario; the card must be equipped.</p>
@@ -570,8 +570,9 @@

 <section class="mb-10">
 	<h3 id="test-script" class="text-lg font-medium mb-2">8.9 Test script</h3>
-	<p class="text-sm mb-2">A <strong>test script</strong> is an ordered list of <strong>actions</strong> and proactive-command <strong>expectations</strong> executed server-side on the equipped card. While a run is active the card belongs to the script: other card actions are refused (HTTP&nbsp;409), and background STATUS polling and the automatic proactive handling are suspended &mdash; a command the card announces stays pending for the script, and both resume when the run ends. The card&rsquo;s own BIP commands (OPEN/SEND/RECEIVE/CLOSE CHANNEL, GET CHANNEL STATUS) are an exception: the terminal answers them for real &mdash; its channels keep working during the run and the BIP events flow for the channels the card itself opened (BIP <em>control</em> stays blocked) &mdash; unless the step&rsquo;s <code class="font-mono text-sm">respond</code> overrides the answer with <code class="font-mono text-sm">raw</code> TLVs or a non-ok result; the run report is kept in memory (also after reloading the page) until <strong>Clear report</strong>. Scripts live in the <strong>server-side store</strong> (<code class="font-mono text-sm">~/.pysim-simple-server/test_scripts.json</code>, <code class="font-mono text-sm">--test-scripts</code>), so every browser manages the same set; edits are written to the store as you change them (there is no Save button &mdash; a draft without a valid step stays local until it is valid); JSON export/import remains for backup and for moving scripts between benches (the old localStorage export imports as-is).</p>
+	<p class="text-sm mb-2">A <strong>test script</strong> is an ordered list of <strong>actions</strong> and proactive-command <strong>expectations</strong> executed server-side on the equipped card. While a run is active the card belongs to the script: other card actions are refused (HTTP&nbsp;409), and background STATUS polling and the automatic proactive handling are suspended &mdash; a command the card announces stays pending for the script, and both resume when the run ends. The card&rsquo;s own BIP commands (OPEN/SEND/RECEIVE/CLOSE CHANNEL, GET CHANNEL STATUS) are an exception: the terminal answers them for real &mdash; its channels keep working during the run and the BIP events flow for the channels the card itself opened (BIP <em>control</em> stays blocked) &mdash; unless the step&rsquo;s <code class="font-mono text-sm">respond</code> overrides the answer with <code class="font-mono text-sm">raw</code> TLVs or a non-ok result; the run report is kept in memory (also after reloading the page) until <strong>Clear report</strong>. Scripts live in the <strong>server-side store</strong> (<code class="font-mono text-sm">~/.pysim-simple-server/test_scripts.json</code>, <code class="font-mono text-sm">--test-scripts</code>), so every browser manages the same set; edits are written to the store as you change them (there is no Save button &mdash; a draft without a valid step stays local until it is valid); JSON export/import remains for backup and for moving scripts between benches (the old localStorage export imports as-is).  A card that stops answering does not hang the run silently: after 30&nbsp;s without progress the watchdog marks the run as failed, naming the step, sets the stop flag and tries to cancel the blocked exchange by disconnecting the PC/SC link on a background thread (the link is marked stale, so the next <strong>Equip</strong> rebuilds it).  A driver with its own card timeout fails the step through the normal path instead.</p>
 	<p class="text-sm mb-2">The pill has <strong>four views</strong>: the <strong>suite list</strong> (default; <em>Add suite</em> / <em>Edit</em> / <em>Run</em> / <em>Export</em> / <em>Delete</em>), the <strong>suite editor</strong> (the suite's scripts in run order &mdash; the optional <em>setup</em> first and the <em>teardown</em> last, marked with role badges; per script <em>Run</em>/<em>Edit</em>/<em>Move</em>/<em>Copy</em>/<em>Delete</em>, the per-member <em>on failure</em> policy <em>stop</em> (default) / <em>continue</em>, and <em>ADM required</em>), the <strong>script editor</strong> (the step list; its <em>Start</em> runs that single script and shows the result under the steps &mdash; not preserved) and the <strong>suite report</strong> (after a suite run: every script's status, its step table with expected/actual, the failing note and a run-log excerpt; the summary with passed/warnings/failed/skipped, wall time, the SCP80 counters before&nbsp;&rarr;&nbsp;after per keyset, the preset and a card-session-changed warning; export as <em>JSON</em> or <em>Markdown</em>).  A <strong>suite run</strong> executes its scripts in order in <strong>one card session, without resets and without resume</strong>; <em>on failure</em> decides whether the suite stops at that script or runs the next one, the <strong>teardown always runs</strong> on a stop or a failure (while the card is in the reader) so the next test starts clean, and a card reset stops the suite and is flagged in the report.  <strong>Start with card reset</strong> resets the card and runs the full re-initialization before the suite starts (the scripts then run against a freshly initialized card); the preset must still match the card's ICCID afterwards and a required ADM is verified again for the new card session.  The initialization consumes the card's initial proactive commands, so a script cannot catch the first command after the reset (they stay in the proactive log).  <strong>ADM required</strong> (on the suite or a script) verifies the matched preset's ADM key <strong>once</strong> before the start &mdash; a failure refuses the run (verify manually with the ADM badge; nothing is retried automatically).  Scripts live in the suite's store (<code class="font-mono text-sm">~/.pysim-simple-server/test_suites.json</code>, <code class="font-mono text-sm">--test-suites</code>); every script belongs to exactly one suite, and a script can be moved or copied to another suite (a bundle export carries the suite plus its scripts; the old script-only export imports into an auto-created &ldquo;Imported scripts&rdquo; suite).</p>
+	<p class="text-sm mb-2">A run can be repeated: the <strong>Iterations</strong> field next to <em>Start</em> / <em>Run suite</em> runs the script or the whole suite that many times &mdash; a <strong>load test</strong>. Successful iterations are not tracked: the status shows only the counter (iteration k / N) and the timing, the per-step log stays empty, and memory use stays flat however high the count. The first iteration that ends in an error stops the loop and keeps its full report (steps, failed checks, run-log excerpt, suite member report); a manual <em>Stop</em> ends it after the current iteration, and a card change between iterations is reported as an error (the next card is never run against another card&rsquo;s script). The count is a run option &mdash; the saved script/suite is unchanged &mdash; and the SCP80 counters continue from what the previous iteration consumed.</p>
 	<ul class="list-disc list-inside text-sm space-y-1 mb-3">
 		<li><strong>Actions</strong> &mdash; <em>Envelope</em> (Event Download: raw event type + data, with an optional source override &mdash; 82 terminal / 83 network) or <em>Event</em> (the semantic forms: the <strong>same form the Phone tab's send dialog uses</strong>, for every modelled event &mdash; the four events the server builds (Location status, Access technology change, Network rejection, Data connection status change) travel as fields it validates and builds at run time, every other event carries the form's built event-data hex (shown live under the form, with the one-ENVELOPE check); the device-identities source comes from the event (e.g. MT call &rarr; network) or its source field), <em>Menu Selection</em> (item id 1&ndash;255, or the item's text with an exact/contains match, case-insensitive by default &mdash; the step editor's <em>Match case</em> choice switches to an exact-case match; the text is resolved against the card's menu, which follows the applet's install parameters), <em>UPDATE/READ file</em> (path + mode; reads verify the file data), <em>raw APDU</em> (no auto-handler), <em>SCP80</em> (C-APDU or a pre-built secured packet, optional keyset number (kvn) / TAR / SPI1 / SPI2 overrides and a command format &mdash; compact, or the TS 102 226 &sect;5.2.1 Command Scripting template <em>expanded</em> (AA) / <em>expanded AE&nbsp;80</em> around the C-APDU; the keys and the counter come from the card preset matched by the equipped card's ICCID &mdash; the run names the stored preset, so it always uses its keysets, TARs and counter &mdash; the server persists the consumed counter into it as the run proceeds), <em>STATUS</em> (with <em>attempts</em> &gt; 1 it polls until the card announces a command) and <em>Proactive drain</em> (poll STATUS and consume every announced command &mdash; the TERMINAL RESPONSE result for each, optionally a different result for the first; the attempt/interval bound; an optional <em>require</em> assertion over what was drained.  An empty poll passes, the final STATUS must be idle, and a still-pending command is handed to the next step &mdash; use it at the start, in the middle or at the end of a script to make sure the card has stopped sending; it also consumes a command a previous step left pending (a step that answered <code class="font-mono text-sm">91XX</code>). Every action has a response check: expected SW and data (exact, or a mask with <code class="font-mono text-sm">?</code> nibble wildcards &mdash; <code class="font-mono text-sm">91??</code>) plus a PoR check for SCP80 (<em>any / no PoR / por_ok</em>, or a custom assertion of the decoded PoR &mdash; status / SW / data; the SEND SHORT MESSAGE transport works the same).</li>
 		<li><strong>Expectations</strong> &mdash; the card must have announced a proactive command in the previous step (SW <code class="font-mono text-sm">91XX</code>, TS 102 221 &sect;7.4.2.1 / TS 102 223 &sect;6.3); a <code class="font-mono text-sm">9000</code> means &ldquo;no command&rdquo;: the step fails at its fail level &mdash; Error stops the script, Warning continues (the tolerant &ldquo;consume a command if there is one&rdquo; form) &mdash; and a STATUS action with <em>attempts</em> &gt; 1 polls when the applet delivers on poll. The command type, qualifier and contents are verified: text / item / raw checks, an <em>Alpha</em> check (the command's Alpha identifier, tag <code class="font-mono text-sm">05</code>/<code class="font-mono text-sm">85</code> &mdash; distinct from the <code class="font-mono text-sm">8D</code> text string), an <em>SMS TPDU</em> check (the submit TPDU of a SEND SHORT MESSAGE: destination / PID / DCS / UDL / user data &mdash; for an applet's own SMS), a <em>PoR</em> check (a SEND SHORT MESSAGE carries the PoR when the packet requested PoR-in-submit, SPI2 bit <code class="font-mono text-sm">20</code>: the response status, the R-APDU SW and the response data; make sure an SCP80 step ran first &mdash; the R-APDU SW lives <em>inside</em> the response payload: an applet's own TAR returns application data, so check its data and leave the SW empty; a refused ENVELOPE (<code class="font-mono text-sm">62XX</code>/<code class="font-mono text-sm">63XX</code>) may carry no response packet at all &mdash; leave the PoR check at <em>any</em> or expect <em>no PoR</em> for such a step) and a <em>Files</em> check (the File List of a REFRESH, e.g. qualifier <code class="font-mono text-sm">00</code> = NAA initialization with full file change notification). The scripted <strong>TERMINAL RESPONSE</strong> is part of the instruction: the <strong>result</strong> offers the full TS 102 223 &sect;8.12.0 list (each entry labelled <code class="font-mono text-sm">&lt;hex&gt; - &lt;alias or name&gt;</code>, plus a manual hex entry; with a command selected the list is grouped by the spec&rsquo;s applicability table (the command-relevant values first, the rest after &mdash; nothing is hidden); the aliases follow the spec &mdash; <code class="font-mono text-sm">ok</code> 00, <code class="font-mono text-sm">modified</code> 07, <code class="font-mono text-sm">cancel</code> 10, <code class="font-mono text-sm">timeout</code> 12, <code class="font-mono text-sm">refused</code> 22, <code class="font-mono text-sm">not_understood</code> 32, &hellip;), the <strong>additional information</strong> byte of the general result (default <code class="font-mono text-sm">00</code>; the &sect;8.12.2 terminal-problem values are suggested), the item identifier, the text (+DCS) and extra TLVs.</li>
@@ -1412,7 +1412,7 @@
 						</div>
 						<label class="flex items-center gap-1 text-sm text-gray-700 dark:text-slate-300"><input id="test-suite-adm" type="checkbox" onchange="testSuiteAdmChanged()"> <span data-l10n="ADM required">ADM required</span></label>
 						<label class="flex items-center gap-1 text-sm text-gray-700 dark:text-slate-300" data-l10n-title="Reset and fully re-initialize the card (ICCID, network state, TERMINAL PROFILE, drain) before the suite starts.  The preset must still match the card's ICCID after the reset; the ADM is verified again for the new card session.  The card's initial proactive commands are consumed by the initialization, so a script cannot catch the first command after the reset (they stay in the proactive log).  A reset later in the run still stops the suite."><input id="test-suite-reset" type="checkbox" onchange="testSuiteResetChanged()"> <span data-l10n="Start with card reset">Start with card reset</span></label>
-						<button id="test-suite-run-btn" data-needs="card" onclick="testSuiteRunCurrent()" class="px-4 py-1.5 text-sm rounded bg-emerald-600 text-white hover:bg-emerald-700 disabled:opacity-40 disabled:cursor-not-allowed" data-l10n="Run suite">Run suite</button>
+						<label class="flex items-center gap-1 text-xs text-gray-600 dark:text-slate-400"><span data-l10n="Iterations">Iterations</span><input id="test-suite-iterations" type="number" min="1" max="1000000" value="1" oninput="testIterationsChanged(this)" class="w-20 border border-gray-300 dark:border-slate-600 text-sm rounded px-2 py-1.5 dark:bg-slate-800"></label> <button id="test-suite-run-btn" data-needs="card" onclick="testSuiteRunCurrent()" class="px-4 py-1.5 text-sm rounded bg-emerald-600 text-white hover:bg-emerald-700 disabled:opacity-40 disabled:cursor-not-allowed" data-l10n="Run suite">Run suite</button>
 						<button onclick="testSuiteExportCurrent()" class="px-3 py-1.5 text-sm rounded bg-gray-200 dark:bg-slate-700 text-gray-700 dark:text-slate-300 hover:bg-gray-300 dark:hover:bg-slate-600" data-l10n="Export">Export</button>
 						<button onclick="testSuiteDeleteCurrent()" class="px-3 py-1.5 text-sm rounded bg-gray-200 dark:bg-slate-700 text-red-600 dark:text-red-400 hover:bg-gray-300 dark:hover:bg-slate-600" data-l10n="Delete">Delete</button>
 					</div>
@@ -1434,7 +1434,7 @@
 							<input id="test-name" oninput="testNameChanged()" class="w-full border border-gray-300 dark:border-slate-600 text-sm rounded px-3 py-1.5 dark:bg-slate-800">
 						</div>
 						<label class="flex items-center gap-1 text-sm text-gray-700 dark:text-slate-300"><input id="test-script-adm" type="checkbox" onchange="testScriptAdmChanged()"> <span data-l10n="ADM required">ADM required</span></label>
-						<button id="test-run-btn" data-needs="card" onclick="testRunStart()" class="px-4 py-1.5 text-sm rounded bg-emerald-600 text-white hover:bg-emerald-700 disabled:opacity-40 disabled:cursor-not-allowed" data-l10n="Start">Start</button>
+						<label class="flex items-center gap-1 text-xs text-gray-600 dark:text-slate-400"><span data-l10n="Iterations">Iterations</span><input id="test-iterations" type="number" min="1" max="1000000" value="1" oninput="testIterationsChanged(this)" class="w-20 border border-gray-300 dark:border-slate-600 text-sm rounded px-2 py-1.5 dark:bg-slate-800"></label> <button id="test-run-btn" data-needs="card" onclick="testRunStart()" class="px-4 py-1.5 text-sm rounded bg-emerald-600 text-white hover:bg-emerald-700 disabled:opacity-40 disabled:cursor-not-allowed" data-l10n="Start">Start</button>
 						<button id="test-stop-btn" onclick="testStop()" class="hidden px-4 py-1.5 text-sm rounded bg-red-600 text-white hover:bg-red-700" data-l10n="Stop">Stop</button>
 						<button onclick="testClearReport()" class="px-3 py-1.5 text-sm rounded bg-gray-200 dark:bg-slate-700 text-gray-700 dark:text-slate-300 hover:bg-gray-300 dark:hover:bg-slate-600" data-l10n="Clear report">Clear report</button>
 					</div>
@@ -1918,7 +1918,22 @@
 			</div>
 			<div class="flex items-center gap-2 text-sm text-gray-600 dark:text-slate-400">
 				<div id="ram-op-spin" class="animate-spin w-4 h-4 border-2 border-blue-500 border-t-transparent rounded-full"></div>
+				<span id="ram-op-insert" class="hidden items-center" aria-hidden="true">
+					<svg viewBox="0 0 72 48" width="64" height="43" class="shrink-0">
+						<!-- the card follows the app's SIM silhouette: a cut corner and the contact pad -->
+						<g class="ram-insert-card text-blue-600 dark:text-blue-400">
+							<path d="M29.5 0.5 H39.5 L45 6 V24 A2.5 2.5 0 0 1 42.5 26.5 H29.5 A2.5 2.5 0 0 1 27 24 V3 A2.5 2.5 0 0 1 29.5 0.5 Z" fill="currentColor"></path>
+							<rect x="31.5" y="16.5" width="9" height="8" rx="1.2" fill="#fff" opacity="0.92"></rect>
+							<path d="M31.5 20.5h9M34.5 16.5v8M37.5 16.5v8" stroke="currentColor" stroke-width="0.55" opacity="0.6"></path>
+						</g>
+						<!-- the reader bar and its slot, drawn after the card so it slides behind -->
+						<rect x="3" y="27" width="66" height="20" rx="4" fill="currentColor" class="text-gray-500 dark:text-slate-400"></rect>
+						<rect x="8" y="33" width="56" height="5" rx="2.5" fill="currentColor" class="text-gray-900" opacity="0.55"></rect>
+						<rect class="ram-insert-glow text-amber-500 dark:text-amber-400" x="8" y="33" width="56" height="5" rx="2.5" fill="none" stroke="currentColor" stroke-width="1.2"></rect>
+					</svg>
+				</span>
 				<span id="ram-op-status" data-l10n="Working...">Working...</span>
+				<button type="button" id="ram-op-sound" onclick="ramBatchToggleSound()" class="ml-auto hidden px-2 py-0.5 text-xs rounded-full border border-gray-300 dark:border-slate-600 text-gray-600 dark:text-slate-300 hover:bg-gray-200 dark:hover:bg-slate-700" aria-pressed="true" title="Sound on" aria-label="Sound on">🔊</button>
 			</div>
 			<div class="mt-2 w-full bg-gray-200 dark:bg-slate-700 rounded h-1.5">
 				<div id="ram-op-bar" class="bg-blue-600 dark:bg-blue-500 h-1.5 rounded" style="width:0%"></div>
@@ -1944,7 +1959,7 @@
 // ===== Version =====
 // Single source of truth for the PWA version: shown in the header and used
 // by the server version check in pysimConnect().
-const SIMPLE_VERSION = '3.35.0';
+const SIMPLE_VERSION = '3.37.0';
 document.getElementById('app-version').textContent = 'v' + SIMPLE_VERSION;

 // ===== view history (the browser Back/Forward buttons) =====
@@ -7516,6 +7531,7 @@ function testViewShow(view) {
 }

 function testRender() {
+	if (!_testIterationsReady) { _testIterationsReady = true; testIterationsInit(); }
 	const views = {suites: 'test-view-suites', suite: 'test-view-suite',
 		script: 'test-view-script', report: 'test-view-report'};
 	Object.keys(views).forEach(v => {
@@ -7527,8 +7543,12 @@ function testRender() {
 	if (banner) {
 		if (running) {
 			const snap = _testRunState || {};
-			const what = snap.suite ? t('Test suite running:') : t('Test script running:');
-			const at = snap.suite ? testSuiteProgress(snap) : ((snap.index + 1) + '/' + snap.total);
+			const load = snap.load && snap.load.enabled ? snap.load : null;
+			const isSuite = !!(snap.suite || (load && load.suite));
+			const what = isSuite ? t('Test suite running:') : t('Test script running:');
+			const at = load
+				? (t('Iteration') + ' ' + (load.iteration || 0) + '/' + load.iterations)
+				: (snap.suite ? testSuiteProgress(snap) : ((snap.index + 1) + '/' + snap.total));
 			banner.textContent = '\u26a0 ' + what + ' ' + (snap.name || '') + ' (' + at + ')';
 			banner.classList.remove('hidden');
 		} else {
@@ -8905,15 +8925,24 @@ function testRenderScriptRun() {
 		stepsEl.innerHTML = '';
 		return;
 	}
+	const load = snap.load && snap.load.enabled ? snap.load : null;
 	let head = esc(snap.name || '');
 	if (snap.status) head += ' · ' + testStatusBadge(snap.status);
 	else if (snap.running) head += ' · ' + esc(t('running'));
+	if (load) {
+		head += ' · ' + esc(t('Iteration')) + ' ' + (load.iteration || 0)
+			+ ' / ' + load.iterations;
+		if (load.done) head += ' · ' + esc(t('done')) + ': ' + load.done;
+	}
 	if (snap.preset) head += ' · ' + esc(t('Preset')) + ': ' + esc(snap.preset);
 	if (snap.scp80_counter) head += ' · counter ' + esc(snap.scp80_counter);
 	if (snap.error) head += ' · ' + esc(snap.error);
 	statusEl.innerHTML = head;
 	statusEl.className = 'text-xs mb-2 text-gray-600 dark:text-slate-300';
-	stepsEl.innerHTML = (snap.steps || []).map(testRunStepHtml).join('');
+	// a load test shows the counter while it runs; only the failure report is
+	// rendered (successful iterations are not tracked)
+	stepsEl.innerHTML = (load && snap.running)
+		? '' : (snap.steps || []).map(testRunStepHtml).join('');
 }

 // ── the suite report ──
@@ -8925,16 +8954,28 @@ function testRenderReport() {
 	if (!titleEl || !sumEl || !memEl) return;
 	const snap = _testRunState;
 	const s = snap && snap.suite;
-	if (!s) {
+	// a load test keeps no suite state on success: the run summary is enough
+	const load = snap && snap.load && snap.load.enabled ? snap.load : null;
+	if (!s && !load) {
 		titleEl.textContent = t('No suite run yet.');
 		sumEl.textContent = '';
 		memEl.innerHTML = '';
 		return;
 	}
-	titleEl.innerHTML = esc(s.name || t('Suite')) + ' · '
+	titleEl.innerHTML = esc((s && s.name) || (snap && snap.name) || t('Suite')) + ' · '
 		+ testStatusBadge(snap.status || (snap.running ? 'running' : ''));
 	const parts = [];
-	const su = s.summary;
+	if (load) {
+		parts.push(t('Iteration') + ' ' + (load.iteration || 0) + ' / ' + load.iterations);
+		if (load.done) parts.push(t('done') + ': ' + load.done);
+		if (load.elapsed_ms !== null && load.elapsed_ms !== undefined) {
+			parts.push((load.elapsed_ms / 1000).toFixed(1) + ' s');
+		}
+		if (load.avg_ms !== null && load.avg_ms !== undefined) {
+			parts.push(t('avg') + ' ' + load.avg_ms + ' ms');
+		}
+	}
+	const su = s && s.summary;
 	if (su) {
 		parts.push(t('passed') + ': ' + su.ok, t('warnings') + ': ' + su.warning,
 			t('failed') + ': ' + su.error, t('skipped') + ': ' + su.skipped);
@@ -8942,19 +8983,19 @@ function testRenderReport() {
 		if (su.wall_ms !== undefined && su.wall_ms !== null) parts.push((su.wall_ms / 1000).toFixed(1) + ' s');
 	}
 	if (snap.preset) parts.push(t('Preset') + ': ' + snap.preset);
-	if (s.start_reset) {
+	if (s && s.start_reset) {
 		const r = s.reset || {};
 		let line = t('started with a card reset');
 		if (r.iccid) line += ' (ICCID ' + r.iccid + ')';
 		if (r.ms !== undefined && r.ms !== null) line += ' \u00b7 ' + (r.ms / 1000).toFixed(1) + ' s';
 		parts.push(line);
 	}
-	if (s.counters_before && Object.keys(s.counters_before).length) {
+	if (s && s.counters_before && Object.keys(s.counters_before).length) {
 		const after = s.counters_after || {};
 		parts.push('SCP80: ' + Object.keys(s.counters_before).map(k =>
 			k + ' ' + (s.counters_before[k] || '-') + ' \u2192 ' + (after[k] || '-')).join(', '));
 	}
-	if (s.session_changed) parts.push('\u26a0 ' + t('card session changed mid-suite'));
+	if (s && s.session_changed) parts.push('\u26a0 ' + t('card session changed mid-suite'));
 	if (snap.adm && snap.adm.required) parts.push('ADM: ' + (snap.adm.verified ? t('verified') : t('not verified')));
 	if (snap.error) parts.push(esc(snap.error));
 	sumEl.innerHTML = esc(parts.join(' · '));
@@ -8962,14 +9003,20 @@ function testRenderReport() {
 	// creates fresh <details>, so an expanded Run log would collapse each
 	// time.  Carry the open state across the re-render (the member list is
 	// fixed at run start, so the index is a stable key).
+	const showMembers = !!s && (!load || snap.status === 'error' || snap.status === 'stopped');
 	const openLogs = {};
+	if (showMembers) {
 		memEl.querySelectorAll('details[data-member]').forEach(d => {
 			openLogs[d.dataset.member] = d.open;
 		});
-	memEl.innerHTML = (s.members || []).map(testReportMemberHtml).join('');
+	}
+	memEl.innerHTML = showMembers
+		? (s.members || []).map(testReportMemberHtml).join('') : '';
+	if (showMembers) {
 		memEl.querySelectorAll('details[data-member]').forEach(d => {
 			if (openLogs[d.dataset.member]) d.open = true;
 		});
+	}
 }

 function testReportMemberHtml(m, i) {
@@ -9003,7 +9050,8 @@ function testDownloadText(name, text, mime) {
 }

 function testReportExportJson() {
-	if (!_testRunState || !_testRunState.suite) {
+	if (!_testRunState || (!_testRunState.suite
+			&& !(_testRunState.load && _testRunState.load.enabled))) {
 		ioStatus('test-io-status', t('No suite run yet.'));
 		return;
 	}
@@ -9011,7 +9059,8 @@ function testReportExportJson() {
 }

 function testReportExportMarkdown() {
-	if (!_testRunState || !_testRunState.suite) {
+	if (!_testRunState || (!_testRunState.suite
+			&& !(_testRunState.load && _testRunState.load.enabled))) {
 		ioStatus('test-io-status', t('No suite run yet.'));
 		return;
 	}
@@ -9117,6 +9166,7 @@ async function testRunStart() {
 	try {
 		const body = {script: {name: cur.name || '', steps: cur.steps,
 			require_adm: !!cur.require_adm}};
+		body.iterations = testIterationsValue('test-iterations');
 		if (cur.id) body.script_id = cur.id;
 		if (presetId) body.preset_id = presetId;
 		_testRunScriptRef = cur;
@@ -9150,6 +9200,7 @@ async function testSuiteRun(i) {
 	}
 	try {
 		const body = {suite_id: s.id};
+		body.iterations = testIterationsValue('test-suite-iterations');
 		if (presetId) body.preset_id = presetId;
 		const resp = await pysimFetch('/api/test/run', body);
 		if (resp.error) { ioStatus('test-io-status', resp.error); return; }
@@ -9164,6 +9215,46 @@ async function testSuiteRun(i) {

 function testSuiteRunCurrent() { testSuiteRun(_testSuiteIdx); }

+// ===== load testing (v3.37.0) =====
+// The iteration count is a run parameter, not part of the stored script or
+// suite: > 1 repeats the run, the successful iterations are discarded and the
+// first failing iteration's report is kept.
+const TEST_ITERATIONS_MAX = 1000000;
+let _testIterationsReady = false;
+
+function testIterationsInit() {
+	let n = 1;
+	try { n = parseInt(localStorage.getItem('simple_test_iterations'), 10) || 1; }
+	catch (e) { /* private mode */ }
+	if (!(n >= 1)) n = 1;
+	if (n > TEST_ITERATIONS_MAX) n = TEST_ITERATIONS_MAX;
+	['test-iterations', 'test-suite-iterations'].forEach(id => {
+		const el = document.getElementById(id);
+		if (el) el.value = String(n);
+	});
+}
+
+function testIterationsChanged(input) {
+	let n = parseInt((input && input.value) || '1', 10);
+	if (!(n >= 1)) n = 1;
+	if (n > TEST_ITERATIONS_MAX) n = TEST_ITERATIONS_MAX;
+	if (input) input.value = String(n);
+	try { localStorage.setItem('simple_test_iterations', String(n)); }
+	catch (e) { /* ignore */ }
+	['test-iterations', 'test-suite-iterations'].forEach(id => {
+		const el = document.getElementById(id);
+		if (el && el !== input) el.value = String(n);
+	});
+}
+
+function testIterationsValue(id) {
+	const el = document.getElementById(id);
+	let n = parseInt((el && el.value) || '1', 10);
+	if (!(n >= 1)) n = 1;
+	if (n > TEST_ITERATIONS_MAX) n = TEST_ITERATIONS_MAX;
+	return n;
+}
+

 async function testStop() {
 	try {
@@ -9254,9 +9345,12 @@ async function testInit() {
 	testRender();
 	try {
 		const st = await pysimFetch('/api/test/status');
-		if (st && st.suite && st.suite.members) {
-			// a suite run (or its finished report): open the report view
-			const idx = testSuites().findIndex(s => s.id === st.suite.id);
+		if (st && ((st.suite && st.suite.members)
+				|| (st.load && st.load.enabled))) {
+			// a suite run (or its finished report, a load run included):
+			// open the report view
+			const idx = st.suite
+				? testSuites().findIndex(s => s.id === st.suite.id) : -1;
 			if (idx >= 0) _testSuiteIdx = idx;
 			_testRunState = st;
 			if (_testView !== 'script') _testView = 'report';
@@ -10273,6 +10367,8 @@ function ramBatchModalOpen() {
 	document.getElementById('ram-op-spin').classList.add('hidden');
 	document.getElementById('ram-op-warn').classList.add('hidden');
 	document.getElementById('ram-op-batch-info').classList.remove('hidden');
+	const sound = document.getElementById('ram-op-sound');
+	if (sound) { sound.classList.remove('hidden'); ramBatchUpdateSoundButton(); }
 	modal.classList.remove('hidden');
 	ramBatchRender();                        // paint from the current state now
 }
@@ -10306,7 +10402,9 @@ function ramBatchModalRender(st, okLabel, actLabel) {
 	document.getElementById('ram-op-bar').style.width = pct + '%';
 	// the counts and the last finished card's verdict
 	const c = st.counts || {};
-	let info = okLabel + ': ' + (c.installed || 0) + ' \u00b7 ' + t('Failed') + ': ' + (c.failed || 0)
+	let info = okLabel + ': ' + (c.installed || 0);
+	if (c.already_absent) info += ' \u00b7 ' + t('Already absent') + ': ' + c.already_absent;
+	info += ' \u00b7 ' + t('Failed') + ': ' + (c.failed || 0)
 		+ ' \u00b7 ' + t('Skipped') + ': ' + (c.skipped || 0);
 	if (st.attempts_total) info += ' \u00b7 ' + st.attempts_total + ' ' + t('attempts');
 	if (st.notice) info += '\n' + st.notice;
@@ -10327,6 +10425,98 @@ function ramBatchModalRender(st, okLabel, actLabel) {
 	document.getElementById('ram-op-close').disabled = active;
 }

+// ===== batch "insert the next card" cue (v3.36.0) =====
+// The waiting phase is the only moment the operator must act, so it gets a
+// card-into-reader animation in the modal, an optional chime (edge-triggered:
+// once per card, not on every 2 s poll) and a tab-title hint while the tab is
+// hidden.  The chime needs an AudioContext unlocked by a gesture: the Start
+// click (and the sound toggle) do that; a reloaded page stays silent until the
+// next click.
+let _ramBatchWaiting = false;
+let _ramBatchAudio = null;
+let _ramBatchTitleBase = '';
+
+function ramBatchSoundEnabled() {
+	try { return localStorage.getItem('simple_ram_batch_sound') !== '0'; }
+	catch (e) { return true; }
+}
+
+function ramBatchUpdateSoundButton() {
+	const btn = document.getElementById('ram-op-sound');
+	if (!btn) return;
+	const on = ramBatchSoundEnabled();
+	btn.textContent = on ? '\u{1F50A}' : '\u{1F507}';
+	btn.setAttribute('aria-pressed', on ? 'true' : 'false');
+	btn.title = t(on ? 'Sound on' : 'Sound off');
+	btn.setAttribute('aria-label', t(on ? 'Sound on' : 'Sound off'));
+}
+
+function ramBatchToggleSound() {
+	try { localStorage.setItem('simple_ram_batch_sound', ramBatchSoundEnabled() ? '0' : '1'); }
+	catch (e) { /* private mode */ }
+	ramBatchUpdateSoundButton();
+	if (ramBatchSoundEnabled()) ramBatchChime();   // a preview when turning it on
+}
+
+function ramBatchAudioUnlock() {
+	try {
+		if (typeof window === 'undefined') return;
+		const Ctor = window.AudioContext || window.webkitAudioContext;
+		if (!Ctor) return;
+		if (!_ramBatchAudio) _ramBatchAudio = new Ctor();
+		if (_ramBatchAudio.state === 'suspended') _ramBatchAudio.resume();
+	} catch (e) { _ramBatchAudio = null; }
+}
+
+function ramBatchChime() {
+	if (!ramBatchSoundEnabled()) return;
+	try {
+		if (typeof window === 'undefined') return;
+		const Ctor = window.AudioContext || window.webkitAudioContext;
+		if (!Ctor) return;
+		if (!_ramBatchAudio) _ramBatchAudio = new Ctor();
+		const ctx = _ramBatchAudio;
+		if (ctx.state === 'suspended') { ctx.resume().catch(() => {}); }
+		const start = ctx.currentTime + 0.02;
+		[880, 660].forEach((freq, i) => {
+			const osc = ctx.createOscillator();
+			const gain = ctx.createGain();
+			osc.type = 'sine';
+			osc.frequency.value = freq;
+			const at = start + i * 0.18;
+			gain.gain.setValueAtTime(0.0001, at);
+			gain.gain.exponentialRampToValueAtTime(0.18, at + 0.02);
+			gain.gain.exponentialRampToValueAtTime(0.0001, at + 0.15);
+			osc.connect(gain);
+			gain.connect(ctx.destination);
+			osc.start(at);
+			osc.stop(at + 0.17);
+		});
+	} catch (e) { /* no audio available: the visual cue remains */ }
+}
+
+function ramBatchTitleHint(waiting) {
+	if (!_ramBatchTitleBase) _ramBatchTitleBase = document.title || 'SIMple';
+	const hint = !!waiting && !!document.hidden;
+	document.title = hint ? t('Insert next card') + ' \u2014 SIMple' : _ramBatchTitleBase;
+}
+
+// Called from ramBatchRender (the 2 s poll): returns true only on the
+// transition into the waiting phase, and updates the animation/title.
+function ramBatchWaitingCue(waiting) {
+	const entered = !!waiting && !_ramBatchWaiting;
+	_ramBatchWaiting = !!waiting;
+	const ins = document.getElementById('ram-op-insert');
+	if (ins) {
+		ins.classList.toggle('hidden', !waiting);
+		ins.classList.toggle('flex', !!waiting);
+	}
+	ramBatchTitleHint(waiting);
+	return entered;
+}
+
+document.addEventListener('visibilitychange', () => ramBatchTitleHint(_ramBatchWaiting));
+
 async function ramBatchFetch() {
 	if (_ramBatchFetching) return;
 	_ramBatchFetching = true;
@@ -10378,6 +10568,10 @@ function ramBatchRender() {
 	const st = ramBatchMergedState();
 	const active = !!st.active;
 	const phase = st.phase || (active ? 'waiting_card' : 'idle');
+	// The waiting cue: the animation in the modal, the chime once per card and
+	// the tab-title hint (v3.36.0).
+	const waiting = active && phase !== 'installing' && phase !== 'stopping';
+	if (ramBatchWaitingCue(waiting)) ramBatchChime();
 	// The success labels follow the operation (v3.29.0): the counts bucket is
 	// 'installed' for every kind - "the operation succeeded".  The last seen
 	// kind is remembered for the modal's title and these labels, so a cleared
@@ -10433,8 +10627,9 @@ function ramBatchRender() {
 	const counts = document.getElementById('ram-batch-counts');
 	if (counts) {
 		const c = st.counts || {};
-		let text = okLabel + ': ' + (c.installed || 0)
-			+ ' \u00b7 ' + t('Failed') + ': ' + (c.failed || 0)
+		let text = okLabel + ': ' + (c.installed || 0);
+		if (c.already_absent) text += ' \u00b7 ' + t('Already absent') + ': ' + c.already_absent;
+		text += ' \u00b7 ' + t('Failed') + ': ' + (c.failed || 0)
 			+ ' \u00b7 ' + t('Skipped') + ': ' + (c.skipped || 0);
 		if (st.attempts_total) text += ' \u00b7 ' + st.attempts_total + ' ' + t('attempts');
 		if (active && _ramBatchProgress && _ramBatchProgress.active && _ramBatchProgress.name) {
@@ -10470,7 +10665,10 @@ function ramBatchAttemptsHtml(attempts) {
 	for (let i = attempts.length - 1; i >= 0; i--) {
 		const a = attempts[i];
 		const when = a.ts ? new Date(a.ts * 1000).toLocaleTimeString() : '';
+		// A delete that found the object already gone is a success, but not a
+		// deletion: say so instead of a plain OK (v3.36.0).
 		const verdict = a.result === 'installed' ? t('OK')
+			: a.result === 'already_absent' ? t('Already absent')
 			: (a.result === 'failed' ? t('Failed') : t('Skipped'));
 		const tone = a.result === 'installed' ? 'text-emerald-600 dark:text-emerald-400'
 			: (a.result === 'failed' ? 'text-red-600 dark:text-red-400' : 'text-amber-600 dark:text-amber-400');
@@ -10526,6 +10724,8 @@ function ramInstallCapFields() {
 }

 async function ramBatchStart() {
+	// The click is the gesture the waiting chime needs later.
+	ramBatchAudioUnlock();
 	// The operation selects the batch kind (v3.29.0): the .cap install, the
 	// applet reinstall (no .cap - the package is loaded) or a GP delete.  The
 	// server builds and validates the plan once at Start.
@@ -10634,6 +10834,10 @@ function ramOpModalBegin(title) {
 		'mt-2 text-xs font-mono whitespace-pre-wrap overflow-auto text-gray-700 dark:text-slate-300';
 	document.getElementById('ram-op-close').disabled = true;
 	document.getElementById('ram-op-spin').classList.remove('hidden');
+	const ins = document.getElementById('ram-op-insert');
+	if (ins) { ins.classList.add('hidden'); ins.classList.remove('flex'); }
+	const snd = document.getElementById('ram-op-sound');
+	if (snd) snd.classList.add('hidden');
 	document.getElementById('ram-op-warn').classList.remove('hidden');
 	document.getElementById('ram-op-batch-info').classList.add('hidden');
 	document.getElementById('ram-op-batch-stop').classList.add('hidden');
@@ -26786,6 +26990,10 @@ const LANG_RU = {
 	'Reinstalling...': 'Переустановка…',
 	'Stopping after the current card...': 'Остановка после текущей карты…',
 	'Insert next card. Waiting for a card...': 'Вставьте следующую карту. Ожидание карты…',
+	'Insert next card': 'Вставьте следующую карту',
+	'Already absent': 'Уже отсутствует',
+	'Sound on': 'Звук включён',
+	'Sound off': 'Звук выключен',
 	'Installed': 'Установлено',
 	'Deleted': 'Удалено',
 	'Reinstalled': 'Переустановлено',
@@ -27083,6 +27291,9 @@ const LANG_RU = {
 	'Move': 'Переместить',
 	'Copy': 'Копировать',
 	'No suite run yet.': 'Набор ещё не запускался.',
+	'Iterations': 'Итерации',
+	'Iteration': 'Итерация',
+	'done': 'выполнено',
 	'Warning': 'Предупреждение',
 	'Text': 'Текст',
 	'id': 'id',
@@ -5,3 +5,27 @@
 .dark .text-heading {
  color: #f1f5f9;
 }
+/* RAM batch: the "insert the next card" cue (v3.36.0).
+   The card slides into the reader slot and back; the slot glows while the
+   run waits.  The static transform is the reduced-motion fallback: the card
+   rests above the slot. */
+.ram-insert-card {
+  /* the resting position is the card just above the slot (0); the insert
+     slide takes it 20px down, behind the reader front */
+  animation: ram-insert 2.8s ease-in-out infinite;
+}
+@keyframes ram-insert {
+  0%, 12% { transform: translateY(0); }
+  45%, 55% { transform: translateY(20px); }
+  88%, 100% { transform: translateY(0); }
+}
+.ram-insert-glow {
+  animation: ram-insert-glow 1.4s ease-in-out infinite;
+}
+@keyframes ram-insert-glow {
+  0%, 100% { opacity: 0.35; }
+  50% { opacity: 0.95; }
+}
+@media (prefers-reduced-motion: reduce) {
+  .ram-insert-card, .ram-insert-glow { animation: none; }
+}
@@ -616,6 +616,10 @@ video {
  pointer-events: none;
 }

+.\!visible {
+  visibility: visible !important;
+}
+
 .visible {
  visibility: visible;
 }
@@ -684,6 +688,11 @@ video {
  grid-column: span 4 / span 4;
 }

+.mx-2 {
+  margin-left: 0.5rem;
+  margin-right: 0.5rem;
+}
+
 .mx-3 {
  margin-left: 0.75rem;
  margin-right: 0.75rem;
@@ -807,6 +816,10 @@ video {
  display: flex;
 }

+.inline-flex {
+  display: inline-flex;
+}
+
 .table {
  display: table;
 }
@@ -843,6 +856,10 @@ video {
  max-height: 8rem;
 }

+.max-h-48 {
+  max-height: 12rem;
+}
+
 .max-h-56 {
  max-height: 14rem;
 }
@@ -935,6 +952,10 @@ video {
  width: 10rem;
 }

+.w-44 {
+  width: 11rem;
+}
+
 .w-48 {
  width: 12rem;
 }
@@ -980,6 +1001,10 @@ video {
  min-width: 0px;
 }

+.min-w-24 {
+  min-width: 6rem;
+}
+
 .min-w-\[12rem\] {
  min-width: 12rem;
 }
@@ -1395,6 +1420,11 @@ video {
  background-color: rgb(219 234 254 / var(--tw-bg-opacity, 1));
 }

+.bg-blue-50 {
+  --tw-bg-opacity: 1;
+  background-color: rgb(239 246 255 / var(--tw-bg-opacity, 1));
+}
+
 .bg-blue-50\/50 {
  background-color: rgb(239 246 255 / 0.5);
 }
@@ -1796,6 +1826,11 @@ video {
  color: rgb(31 41 55 / var(--tw-text-opacity, 1));
 }

+.text-gray-900 {
+  --tw-text-opacity: 1;
+  color: rgb(17 24 39 / var(--tw-text-opacity, 1));
+}
+
 .text-green-600 {
  --tw-text-opacity: 1;
  color: rgb(22 163 74 / var(--tw-text-opacity, 1));
@@ -1893,10 +1928,61 @@ video {
  filter: var(--tw-blur) var(--tw-brightness) var(--tw-contrast) var(--tw-grayscale) var(--tw-hue-rotate) var(--tw-invert) var(--tw-saturate) var(--tw-sepia) var(--tw-drop-shadow);
 }

+.transition {
+  transition-property: color, background-color, border-color, text-decoration-color, fill, stroke, opacity, box-shadow, transform, filter, backdrop-filter;
+  transition-timing-function: cubic-bezier(0.4, 0, 0.2, 1);
+  transition-duration: 150ms;
+}
+
 .dark .text-heading {
  color: #f1f5f9;
 }

+/* RAM batch: the "insert the next card" cue (v3.36.0).
+   The card slides into the reader slot and back; the slot glows while the
+   run waits.  The static transform is the reduced-motion fallback: the card
+   rests above the slot. */
+
+.ram-insert-card {
+  /* the resting position is the card just above the slot (0); the insert
+     slide takes it 20px down, behind the reader front */
+  animation: ram-insert 2.8s ease-in-out infinite;
+}
+
+@keyframes ram-insert {
+  0%, 12% {
+    transform: translateY(0);
+  }
+
+  45%, 55% {
+    transform: translateY(20px);
+  }
+
+  88%, 100% {
+    transform: translateY(0);
+  }
+}
+
+.ram-insert-glow {
+  animation: ram-insert-glow 1.4s ease-in-out infinite;
+}
+
+@keyframes ram-insert-glow {
+  0%, 100% {
+    opacity: 0.35;
+  }
+
+  50% {
+    opacity: 0.95;
+  }
+}
+
+@media (prefers-reduced-motion: reduce) {
+  .ram-insert-card, .ram-insert-glow {
+    animation: none;
+  }
+}
+
 .hover\:bg-amber-100:hover {
  --tw-bg-opacity: 1;
  background-color: rgb(254 243 199 / var(--tw-bg-opacity, 1));
@@ -1956,11 +2042,6 @@ video {
  background-color: rgb(55 65 81 / var(--tw-bg-opacity, 1));
 }

-.hover\:bg-red-100:hover {
-  --tw-bg-opacity: 1;
-  background-color: rgb(254 226 226 / var(--tw-bg-opacity, 1));
-}
-
 .hover\:bg-red-200:hover {
  --tw-bg-opacity: 1;
  background-color: rgb(254 202 202 / var(--tw-bg-opacity, 1));
@@ -2093,6 +2174,10 @@ video {
  background-color: rgb(30 58 138 / 0.1);
 }

+.dark\:bg-blue-900\/20:is(.dark *) {
+  background-color: rgb(30 58 138 / 0.2);
+}
+
 .dark\:bg-blue-900\/40:is(.dark *) {
  background-color: rgb(30 58 138 / 0.4);
 }
@@ -2258,10 +2343,6 @@ video {
  background-color: rgb(6 78 59 / 0.4);
 }

-.dark\:hover\:bg-red-900\/40:hover:is(.dark *) {
-  background-color: rgb(127 29 29 / 0.4);
-}
-
 .dark\:hover\:bg-red-900\/50:hover:is(.dark *) {
  background-color: rgb(127 29 29 / 0.5);
 }
@@ -2289,6 +2370,11 @@ video {
  color: rgb(209 213 219 / var(--tw-text-opacity, 1));
 }

+.dark\:hover\:text-slate-200:hover:is(.dark *) {
+  --tw-text-opacity: 1;
+  color: rgb(226 232 240 / var(--tw-text-opacity, 1));
+}
+
 .dark\:hover\:text-slate-300:hover:is(.dark *) {
  --tw-text-opacity: 1;
  color: rgb(203 213 225 / var(--tw-text-opacity, 1));
@@ -2328,7 +2414,8 @@ video {
  .lg\:flex-row {
    flex-direction: row;
  }
-}/* ===== Theme contrast adjustments ===== */
+}
+/* ===== Theme contrast adjustments ===== */

 /* Light theme: darken softer (muted) text one step for higher contrast */
 .text-gray-400 {
@@ -1,4 +1,4 @@
-const CACHE = 'simple-v469';
+const CACHE = 'simple-v477';
 const URLS = [
  'index.html',
  'help.html',
@@ -28,7 +28,10 @@ for (const fn of ['ramIsBatchOp', 'ramBatchPanelVisible', 'ramBatchMergedState',
 	'ramBatchStepsText',
 	'ramBatchAttemptsHtml', 'ramBatchRender', 'ramBatchApplyStatus', 'ramBatchFetch',
 	'ramBatchStart', 'ramBatchStop', 'ramBatchClear', 'ramOpProgressText',
-	'ramOpModalClose', 'esc']) {
+	'ramOpModalClose', 'esc',
+	'ramBatchSoundEnabled', 'ramBatchUpdateSoundButton', 'ramBatchToggleSound',
+	'ramBatchAudioUnlock', 'ramBatchChime', 'ramBatchTitleHint',
+	'ramBatchWaitingCue']) {
 	code += extractFunc(html, fn, ASYNC_FNS.indexOf(fn) >= 0) + '\n';
 }

@@ -39,7 +42,9 @@ const els = {};
 function setEl(id, value) {
 	els[id] = {
 		id: id, value: value === undefined ? '' : value, textContent: '', innerHTML: '',
-		disabled: false, style: {},
+		disabled: false, style: {}, title: '', attrs: {},
+		setAttribute(k, v) { els[id].attrs[k] = String(v); },
+		getAttribute(k) { return els[id].attrs[k]; },
 		classes: {},
 		classList: {
 			add(c) { els[id].classes[c] = true; },
@@ -64,7 +69,8 @@ function clearEls() {
 function setModalEls() {
 	for (const id of ['ram-op-modal', 'ram-op-title', 'ram-op-status', 'ram-op-bar',
 		'ram-op-steps', 'ram-op-spin', 'ram-op-warn', 'ram-op-close',
-		'ram-op-batch-info', 'ram-op-batch-stop', 'ram-op-batch-clear']) setEl(id, '');
+		'ram-op-batch-info', 'ram-op-batch-stop', 'ram-op-batch-clear',
+		'ram-op-insert', 'ram-op-sound']) setEl(id, '');
 }
 function setPanelEls() {
 	for (const id of ['ram-op', 'ram-batch-panel', 'ram-batch-start', 'ram-batch-stop',
@@ -75,7 +81,17 @@ function setPanelEls() {

 const app = new Function('els', `
 	var sandbox = { fetched: [], replies: {}, fetchError: null, alerts: [], visibility: true };
-	var document = { getElementById: id => els[id] || null };
+	var document = {
+		getElementById: id => els[id] || null,
+		hidden: false, title: 'SIMple',
+		addEventListener: () => {},
+	};
+	var _ls = {};
+	var localStorage = {
+		getItem: k => (k in _ls ? _ls[k] : null),
+		setItem: (k, v) => { _ls[k] = String(v); },
+		removeItem: k => { delete _ls[k]; },
+	};
 	var t = k => k;
 	var isViewVisible = () => sandbox.visibility;
 	var alert = m => sandbox.alerts.push(String(m));
@@ -109,6 +125,9 @@ const app = new Function('els', `
 	var _ramBatchFetchAt = 0;
 	var _ramBatchSummary = null;
 	var _ramBatchProgress = null;
+	var _ramBatchWaiting = false;
+	var _ramBatchAudio = null;
+	var _ramBatchTitleBase = '';
 	var RAM_BATCH_FETCH_STALE_MS = ${RAM_BATCH_FETCH_STALE_MS};
 	${code}
 	sandbox.getLast = () => _ramBatchLast;
@@ -134,6 +153,16 @@ const app = new Function('els', `
 	sandbox.ramBatchStart = ramBatchStart;
 	sandbox.ramBatchStop = ramBatchStop;
 	sandbox.ramBatchClear = ramBatchClear;
+	sandbox.ramBatchWaitingCue = ramBatchWaitingCue;
+	sandbox.ramBatchTitleHint = ramBatchTitleHint;
+	sandbox.ramBatchSoundEnabled = ramBatchSoundEnabled;
+	sandbox.ramBatchUpdateSoundButton = ramBatchUpdateSoundButton;
+	sandbox.ramBatchToggleSound = ramBatchToggleSound;
+	sandbox.setHidden = v => { document.hidden = !!v; };
+	sandbox.getTitle = () => document.title;
+	sandbox.resetWaiting = () => { _ramBatchWaiting = false; };
+	sandbox.storage = _ls;
+	sandbox.resetStorage = () => { for (const k of Object.keys(_ls)) delete _ls[k]; };
 	return sandbox;
 `)(els);

@@ -639,3 +668,96 @@ test('the stop and clear calls hit their endpoints', async () => {
 	await new Promise(r => setImmediate(r));
 	assert.deepStrictEqual(app.fetched, ['/api/ram-batch/stop', '/api/ram-batch/clear', '/api/ram-batch/status']);
 });
+
+
+test('the waiting cue fires once per card and drives the insert animation (v3.36.0)', () => {
+	clearEls();
+	setPanelEls();
+	setModalEls();
+	app.resetWaiting();
+	assert.strictEqual(app.ramBatchWaitingCue(true), true, 'entering waiting cues');
+	assert.strictEqual(app.ramBatchWaitingCue(true), false, 'still waiting: no repeat');
+	assert.strictEqual(app.ramBatchWaitingCue(false), false, 'installing resets');
+	assert.strictEqual(app.ramBatchWaitingCue(true), true, 'the next card cues again');
+	assert.ok(els['ram-op-insert'].classes['flex'] === true);
+	assert.ok(!els['ram-op-insert'].classes['hidden']);
+	app.ramBatchWaitingCue(false);
+	assert.ok(els['ram-op-insert'].classes['hidden']);
+	// the render drives it from the phase
+	app.setLast({ active: true, phase: 'waiting_card', attempts: [], counts: {}, attempts_total: 0, notice: '' });
+	app.ramBatchRender();
+	assert.ok(!els['ram-op-insert'].classes['hidden']);
+	app.setLast({ active: true, phase: 'installing', current: { iccid: '8970119000004600098' },
+		attempts: [], counts: {}, attempts_total: 1 });
+	app.ramBatchRender();
+	assert.ok(els['ram-op-insert'].classes['hidden']);
+});
+
+test('the sound toggle persists and updates the button (v3.36.0)', () => {
+	clearEls();
+	setModalEls();
+	app.resetStorage();
+	assert.strictEqual(app.ramBatchSoundEnabled(), true, 'sound is on by default');
+	app.ramBatchUpdateSoundButton();
+	assert.strictEqual(els['ram-op-sound'].textContent, '\u{1F50A}');
+	assert.strictEqual(els['ram-op-sound'].attrs['aria-pressed'], 'true');
+	app.ramBatchToggleSound();
+	assert.strictEqual(app.ramBatchSoundEnabled(), false);
+	assert.strictEqual(app.storage['simple_ram_batch_sound'], '0');
+	assert.strictEqual(els['ram-op-sound'].textContent, '\u{1F507}');
+	assert.strictEqual(els['ram-op-sound'].attrs['aria-pressed'], 'false');
+	app.ramBatchToggleSound();
+	assert.strictEqual(app.ramBatchSoundEnabled(), true);
+});
+
+test('the tab title hints while the run waits and the tab is hidden (v3.36.0)', () => {
+	app.setHidden(false);
+	app.ramBatchTitleHint(true);
+	assert.strictEqual(app.getTitle(), 'SIMple');
+	app.setHidden(true);
+	app.ramBatchTitleHint(true);
+	assert.strictEqual(app.getTitle(), 'Insert next card \u2014 SIMple');
+	app.setHidden(false);
+	app.ramBatchTitleHint(true);
+	assert.strictEqual(app.getTitle(), 'SIMple');
+});
+
+test('the waiting cue is wired into the poll render, the Start gesture and the CSS (v3.36.0)', () => {
+	const render = extractFunc(html, 'ramBatchRender');
+	assert.match(render, /const waiting = active && phase !== 'installing' && phase !== 'stopping';/);
+	assert.match(render, /if \(ramBatchWaitingCue\(waiting\)\) ramBatchChime\(\);/);
+	assert.match(extractFunc(html, 'ramBatchStart'), /ramBatchAudioUnlock\(\)/);
+	assert.match(html, /id="ram-op-insert"/);
+	assert.match(html, /id="ram-op-sound"/);
+	const css = fs.readFileSync(path.join(__dirname, '..', 'style.css'), 'utf8');
+	assert.match(css, /\.ram-insert-card/);
+	assert.match(css, /@keyframes ram-insert/);
+	assert.match(css, /\.ram-insert-glow/);
+	assert.match(css, /prefers-reduced-motion: reduce/);
+});
+
+test('an already-absent delete is reported as such, not as a deletion (v3.36.0)', () => {
+	clearEls();
+	setPanelEls();
+	setModalEls();
+	app.resetBatchModal();
+	app.setLast({ active: false, phase: 'finished', kind: 'delete',
+		attempts: [{
+			iccid: '8970119000000115026', preset: 'Alfa', result: 'already_absent',
+			steps: [{ name: 'DELETE (application)', sw: '6A88', ok: true, not_found: true }],
+			steps_total: 1, duration_ms: 1200,
+		}],
+		counts: { installed: 2, already_absent: 1, failed: 0, skipped: 0 },
+		attempts_total: 3, notice: 'Card 8970119000000115026 was already absent.' });
+	app.ramBatchRender();
+	// the counts separate a real deletion from a no-op
+	assert.match(els['ram-batch-counts'].textContent, /Deleted: 2/);
+	assert.match(els['ram-batch-counts'].textContent, /Already absent: 1/);
+	// the attempt's verdict says it - not a plain OK - and keeps the step line
+	assert.match(els['ram-batch-attempts'].innerHTML, /Already absent/);
+	assert.match(els['ram-batch-attempts'].innerHTML, /6A88 \(not found\)/);
+	// the modal carries the same counts
+	app.ramBatchModalOpen();
+	assert.match(els['ram-op-batch-info'].textContent, /Deleted: 2/);
+	assert.match(els['ram-op-batch-info'].textContent, /Already absent: 1/);
+});
@@ -0,0 +1,166 @@
+const { test } = require('node:test');
+const assert = require('node:assert');
+const fs = require('node:fs');
+const path = require('node:path');
+
+const html = fs.readFileSync(path.join(__dirname, '..', 'index.html'), 'utf8');
+
+function extractFunc(src, name) {
+	const re = new RegExp('function\\s+' + name + '\\s*\\([^)]*\\)\\s*\\{');
+	const m = re.exec(src);
+	if (!m) throw new Error('function ' + name + ' not found');
+	let i = m.index + m[0].length - 1;
+	let depth = 0;
+	for (; i < src.length; i++) {
+		if (src[i] === '{') depth++;
+		else if (src[i] === '}') {
+			depth--;
+			if (depth === 0) break;
+		}
+	}
+	return src.slice(m.index, i + 1);
+}
+
+const FNS = ['esc', 'testStatusBadge', 'testRenderScriptRun', 'testRenderReport',
+	'testIterationsInit', 'testIterationsChanged', 'testIterationsValue'];
+const code = FNS.map(f => extractFunc(html, f)).join('\n');
+
+function fakeEl(id) {
+	return { id, innerHTML: '', textContent: '', className: '', value: '',
+		querySelectorAll: () => [] };
+}
+
+function build(els, storage, curRef) {
+	const fn = new Function('els', 'storage', 'curRef', `
+		var _testRunState = null;
+		var _testRunScriptRef = curRef.value;
+		var TEST_ITERATIONS_MAX = 1000000;
+		var _testIterationsReady = false;
+		var localStorage = storage;
+		var document = { getElementById: id => els[id] || null };
+		var t = k => k;
+		function testCurrentScript() { return curRef.value; }
+		function testRunStepHtml(s) { return '<STEP ' + (s.note || '') + '>'; }
+		function testReportMemberHtml(m, i) { return '<MEMBER ' + (m.name || '') + '>'; }
+		${code}
+		return {
+			setState: v => { _testRunState = v; },
+			getState: () => _testRunState,
+			testRenderScriptRun, testRenderReport,
+			testIterationsValue, testIterationsChanged, testIterationsInit,
+		};
+	`);
+	return fn(els, storage, curRef);
+}
+
+test('the iteration count clamps and persists (v3.37.0)', () => {
+	const els = { 'test-iterations': fakeEl('test-iterations'),
+		'test-suite-iterations': fakeEl('test-suite-iterations') };
+	const storage = { data: {}, getItem(k) { return this.data[k] === undefined ? null : this.data[k]; },
+		setItem(k, v) { this.data[k] = String(v); } };
+	const app = build(els, storage, { value: null });
+
+	els['test-iterations'].value = '5';
+	assert.strictEqual(app.testIterationsValue('test-iterations'), 5);
+	assert.strictEqual(app.testIterationsValue('missing'), 1);
+	els['test-iterations'].value = '99999999';
+	assert.strictEqual(app.testIterationsValue('test-iterations'), 1000000);
+	els['test-iterations'].value = 'nope';
+	assert.strictEqual(app.testIterationsValue('test-iterations'), 1);
+
+	app.testIterationsChanged(els['test-iterations']);
+	els['test-iterations'].value = '7';
+	app.testIterationsChanged(els['test-iterations']);
+	assert.strictEqual(storage.data.simple_test_iterations, '7');
+	assert.strictEqual(els['test-suite-iterations'].value, '7', 'the other input follows');
+
+	// init restores the remembered value
+	storage.data.simple_test_iterations = '12';
+	app.testIterationsInit();
+	assert.strictEqual(els['test-iterations'].value, '12');
+	assert.strictEqual(els['test-suite-iterations'].value, '12');
+});
+
+test('the run bodies carry the iterations (v3.37.0)', () => {
+	const script = extractFunc(html, 'testRunStart');
+	assert.match(script, /body\.iterations = testIterationsValue\('test-iterations'\)/);
+	const suite = extractFunc(html, 'testSuiteRun');
+	assert.match(suite, /body\.iterations = testIterationsValue\('test-suite-iterations'\)/);
+	// the inputs live next to the two run buttons
+	assert.match(html, /id="test-iterations" type="number" min="1" max="1000000"/);
+	assert.match(html, /id="test-suite-iterations" type="number" min="1" max="1000000"/);
+});
+
+test('a running load test shows the counter, not the steps (v3.37.0)', () => {
+	const els = { 'test-run-status': fakeEl('test-run-status'),
+		'test-run-steps': fakeEl('test-run-steps') };
+	const curRef = { value: { id: 's1' } };
+	const app = build(els, { getItem: () => null, setItem: () => {} }, curRef);
+	app.setState({ running: true, script_id: 's1', name: 'S', steps: [],
+		load: { enabled: true, iterations: 100, iteration: 3, done: 2 } });
+	app.testRenderScriptRun();
+	assert.match(els['test-run-status'].innerHTML, /Iteration 3 \/ 100/);
+	assert.match(els['test-run-status'].innerHTML, /done: 2/);
+	assert.strictEqual(els['test-run-steps'].innerHTML, '', 'steps are not tracked');
+});
+
+test('a load-test failure shows the failing iteration steps (v3.37.0)', () => {
+	const els = { 'test-run-status': fakeEl('test-run-status'),
+		'test-run-steps': fakeEl('test-run-steps') };
+	const curRef = { value: { id: 's1' } };
+	const app = build(els, { getItem: () => null, setItem: () => {} }, curRef);
+	app.setState({ running: false, status: 'error', script_id: 's1', name: 'S',
+		steps: [{ note: 'boom' }], error: 'iteration 4 failed',
+		load: { enabled: true, iterations: 100, iteration: 4, done: 3, failed_iteration: 4 } });
+	app.testRenderScriptRun();
+	assert.match(els['test-run-status'].innerHTML, /Iteration 4 \/ 100/);
+	assert.match(els['test-run-status'].innerHTML, /iteration 4 failed/);
+	assert.match(els['test-run-steps'].innerHTML, /<STEP boom>/);
+});
+
+test('the suite report summarises a load run and hides successful iterations (v3.37.0)', () => {
+	const els = { 'test-report-title': fakeEl('test-report-title'),
+		'test-report-summary': fakeEl('test-report-summary'),
+		'test-report-members': fakeEl('test-report-members') };
+	const app = build(els, { getItem: () => null, setItem: () => {} }, { value: null });
+	// a completed load run: no suite state, only the summary
+	app.setState({ running: false, status: 'ok', name: 'Load suite',
+		load: { enabled: true, iterations: 50, iteration: 50, done: 50,
+			elapsed_ms: 12000, avg_ms: 240, failed_iteration: null }, suite: null });
+	app.testRenderReport();
+	assert.match(els['test-report-title'].innerHTML, /Load suite/);
+	assert.match(els['test-report-summary'].innerHTML, /Iteration 50 \/ 50/);
+	assert.match(els['test-report-summary'].innerHTML, /done: 50/);
+	assert.strictEqual(els['test-report-members'].innerHTML, '');
+	// a failure keeps that iteration's members
+	app.setState({ running: false, status: 'error', name: 'Load suite',
+		error: 'iteration 7 failed',
+		load: { enabled: true, iterations: 50, iteration: 7, done: 6,
+			elapsed_ms: 2000, avg_ms: 300, failed_iteration: 7 },
+		suite: { name: 'Load suite', summary: { ok: 0, warning: 0, error: 1, skipped: 0, stopped: false },
+			members: [{ name: 'A', status: 'error' }] } });
+	app.testRenderReport();
+	assert.match(els['test-report-summary'].innerHTML, /Iteration 7 \/ 50/);
+	assert.match(els['test-report-members'].innerHTML, /<MEMBER A>/);
+	// while running, the members are not shown (results are not tracked)
+	app.setState({ running: true, status: null, name: 'Load suite',
+		load: { enabled: true, iterations: 50, iteration: 2, done: 1 },
+		suite: { name: 'Load suite', summary: null, members: [{ name: 'A', status: 'running' }] } });
+	app.testRenderReport();
+	assert.strictEqual(els['test-report-members'].innerHTML, '');
+});
+
+test('a successful load run can be exported and reopened after a reload (v3.37.0)', () => {
+	// the report is renderable with suite: null, so the export buttons and the
+	// reload path must accept a load run too
+	for (const fn of ['testReportExportJson', 'testReportExportMarkdown']) {
+		const src = extractFunc(html, fn);
+		assert.match(src, /_testRunState\.load && _testRunState\.load\.enabled/);
+	}
+	const init = extractFunc(html, 'testInit');
+	assert.match(init, /st\.load && st\.load\.enabled/);
+	assert.match(init, /const idx = st\.suite\s*\?/, 'a null suite must not break the reopen');
+	// the banner knows a load suite even when its state is cleared
+	const render = extractFunc(html, 'testRender');
+	assert.match(render, /const isSuite = !!\(snap\.suite \|\| \(load && load\.suite\)\);/);
+});
@@ -52,6 +52,9 @@ eval(extractFunc(html, 'testScriptProblem'));
 eval(extractFunc(html, 'testPorSpecSummary'));
 eval(extractFunc(html, 'testCheckSummary'));
 eval(extractFunc(html, 'testStepSummary'));
+// the load-test run parameter (v3.37.0): the run bodies read the count
+eval(html.match(/const TEST_ITERATIONS_MAX = \d+;/)[0].replace('const ', 'var '));
+eval(extractFunc(html, 'testIterationsValue'));
 eval(extractFunc(html, 'testCommandOptions'));
 globalThis.t = s => s;
 globalThis.esc = s => String(s);
@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"

 [project]
 name = "pysim-simple-server"
-version = "3.35.0"
+version = "3.37.0"
 description = "HTTP REST server wrapping pysim for the SIMple PWA"
 license = { text = "GPL-2.0-or-later" }
 requires-python = ">=3.9"
@@ -37,7 +37,7 @@ from osmocom.tlv import BER_TLV_IE
 from osmocom.utils import rpad


-VERSION = '3.35.0'
+VERSION = '3.37.0'

 MAX_ENVELOPE_SEGMENTS = 5  # max SMS segments for outgoing C-APDU in ENVELOPE

@@ -2914,7 +2914,7 @@ def _ram_batch_new_state(plan):
    return {'active': True, 'phase': 'waiting_card', 'stop': False,
            'started': time.time(), 'stopped': None, 'plan': plan,
            'session': None, 'seen': {}, 'attempts': [], 'attempts_total': 0,
-            'counts': {'installed': 0, 'failed': 0, 'skipped': 0},
+            'counts': {'installed': 0, 'already_absent': 0, 'failed': 0, 'skipped': 0},
            'current': None, 'notice': ''}


@@ -2939,6 +2939,8 @@ def _ram_batch_record(state, rec):
            verb = {'delete': 'deleted',
                    'reinstall-app': 'reinstalled'}.get(kind, 'installed')
            state['notice'] = 'Card %s %s.' % (who, verb)
+        elif result == 'already_absent':
+            state['notice'] = 'Card %s was already absent.' % who
        elif result == 'failed':
            step = rec.get('failed_step') or ''
            sw = rec.get('por_sw') or rec.get('sw') or ''
@@ -2947,8 +2949,9 @@ def _ram_batch_record(state, rec):
        else:
            state['notice'] = 'Card %s skipped: %s.' % (who, rec.get('reason') or '')
    detail = rec.get('reason') or rec.get('failed_step') or ''
+    label = {'already_absent': 'already absent'}.get(result, result)
    sys.stderr.write('RAM-BATCH: %s%s%s\n' % (
-        result, (' ' + who) if rec.get('iccid') else '', (' - ' + detail) if detail else ''))
+        label, (' ' + who) if rec.get('iccid') else '', (' - ' + detail) if detail else ''))


 def _ram_batch_failure_fields(result):
@@ -3044,6 +3047,15 @@ def _ram_batch_handle_card(state, server):
               'module_aid': result.get('module_aid'),
               'nv_delta': result.get('nv_delta')}
        rec.update(_ram_batch_steps_fields(result))
+        # A delete run whose DELETE answered 'not found' is a success (the
+        # object is gone, the goal is met) but not a deletion: report it as
+        # "already absent" and count it separately (v3.36.0).  Only the delete
+        # kind: a reinstall's tolerated pre-delete is not the outcome.
+        if (state.get('plan') or {}).get('kind') == 'delete' and any(
+                s.get('delete_not_found') for s in (result.get('steps') or [])):
+            outcome = 'already_absent'
+            rec['result'] = outcome
+            rec['not_found'] = True
    elif result is not None:
        outcome = 'failed'
        rec = {'iccid': iccid, 'preset': name, 'result': outcome,
@@ -3178,7 +3190,7 @@ def _ram_batch_status():
        if _RAM_BATCH is None:
            return {'active': False, 'phase': 'idle', 'attempts': [],
                    'attempts_total': 0, 'current': None, 'notice': '',
-                    'counts': {'installed': 0, 'failed': 0, 'skipped': 0},
+                    'counts': {'installed': 0, 'already_absent': 0, 'failed': 0, 'skipped': 0},
                    'started': None, 'stopped': None, 'progress': None,
                    'kind': None}
        return {'active': bool(_RAM_BATCH.get('active')),
@@ -3580,9 +3592,12 @@ PLI_QUALIFIER_NAMES = {
 # specs define no contents for them ("reserved for 3GPP/3GPP2") and a
 # fabricated object would mislead the application more than no answer.
 PLI_DEFAULTS = {
-    # Location Information (8.19): PLMN 001/01 (the test network, TS 23.003)
-    # + LAC 0000
-    0x00: '1305' + '00F110' + '0000',
+    # Location Information (TS 31.111 8.19.1, the GERAN form a PLI answer
+    # uses): PLMN 001/01 (the test network, TS 23.003) + LAC 0001 + Cell ID
+    # 0001.  Length 05 (no Cell ID) is the *Network Rejection event* form -
+    # a PLI answer must carry the Cell ID (length 07), otherwise the location
+    # reads as empty and a card applet may re-request it forever.
+    0x00: '1307' + '00F110' + '0001' + '0001',
    # IMEI (8.20): the classic test IMEI 490154203237518, nibble-swapped
    # BCD with the odd 15th digit padded by F (TS 124 008 Mobile Identity)
    0x01: '1408' + '94104502237315F8',
@@ -6596,15 +6611,18 @@ def _parse_setup_menu_items(raw):
    return _attach_nai(items, nai)


-def _handle_proactive_chain(scc, sw91, on_fetch=None, status_poll=True, max_rounds=None):
+def _handle_proactive_chain(scc, sw91, on_fetch=None, status_poll=True,
+                            max_rounds=_POLL_CHAIN_MAX):
    """Run a FETCH/TERMINAL RESPONSE chain; marks the card as busy so that
    terminal-initiated ENVELOPEs (Data available, Channel status, timers) wait.
    ``status_poll=False`` skips the trailing STATUS (used by the profile
    switch, where the card is mid-switch and must not be queried further).
-    ``max_rounds`` bounds the answers per call (the background poll passes
-    _POLL_CHAIN_MAX so a card that re-queues a command cannot hold the card
-    lock forever); ``_TEST_RUNNING`` aborts the chain before the next FETCH -
-    a test run owns the card and the pending command stays for the script."""
+    ``max_rounds`` bounds the answers per call - the default is
+    _POLL_CHAIN_MAX, so a card that re-queues a command (a chatty applet)
+    cannot hold the card lock forever from *any* call site; a caller may pass
+    a different bound.  ``_TEST_RUNNING`` aborts the chain before the next
+    FETCH - a test run owns the card and the pending command stays for the
+    script."""
    global _PROACTIVE_BUSY
    _PROACTIVE_BUSY = True
    try:
@@ -7044,9 +7062,19 @@ _TEST_RUN = {
    'preset': None, 'script_id': None,
    'scp80_counter': None, 'scp80_counters': {},
    'started': None, 'finished': None, 'error': None,
+    # the stuck-card watchdog (v3.37.0): `heartbeat` is the last runner
+    # progress mark; `stuck` makes the final status an error even when the
+    # stop flag ended the loop
+    'stuck': False, 'heartbeat': None,
    # suite runs (v3.22.0): None for a single script; the members list and the
    # report summary live here, plus the run-log buffer/adm state
    'suite': None, 'log': [], 'log_prefix': '', 'adm': None,
+    # load testing (v3.37.0): a run parameter, not a stored option.  `done`
+    # counts the successful iterations (their results are discarded); the
+    # failing iteration's steps/log stay as the report.
+    'load': {'enabled': False, 'iterations': None, 'iteration': 0, 'done': 0,
+             'failed_iteration': None, 'started': None, 'finished': None,
+             'elapsed_ms': None, 'avg_ms': None},
 }
 _TEST_LOCK = threading.Lock()
 _TEST_THREAD = None
@@ -7136,11 +7164,158 @@ def _test_command_type(name):
    return _TEST_COMMAND_TYPES.get((name or '').upper())


+_TEST_LOAD_MAX = 1000000
+
+
+def _test_load_state(iterations, suite=False):
+    """The `load` block of a run: disabled for a normal single run.  `suite`
+    lets the PWA label a suite load test even after its state is cleared."""
+    enabled = iterations > 1
+    return {'enabled': enabled, 'suite': bool(suite),
+            'iterations': iterations if enabled else None,
+            'iteration': 0, 'done': 0, 'failed_iteration': None,
+            'started': time.time() if enabled else None, 'finished': None,
+            'elapsed_ms': None, 'avg_ms': None}
+
+
+def _test_load_iterations(body):
+    """The run's iteration count (1 when omitted).  Raises ValueError for a
+    bad value (the endpoint answers 400)."""
+    raw = (body or {}).get('iterations')
+    if raw is None or raw == '':
+        return 1
+    try:
+        n = int(raw)
+    except (TypeError, ValueError):
+        raise ValueError('iterations must be an integer')
+    if isinstance(raw, float) and raw != n:
+        raise ValueError('iterations must be an integer')
+    if n < 1 or n > _TEST_LOAD_MAX:
+        raise ValueError('iterations must be between 1 and %d' % _TEST_LOAD_MAX)
+    return n
+
+
 def _test_state_snapshot():
    with _TEST_LOCK:
        return json.loads(json.dumps(_TEST_RUN))


+# ─── the stuck-card watchdog (v3.37.0) ───────────────────────────────────
+# pyscard's SCardTransmit is blocking and has no timeout: a card that stops
+# answering (no SW, no driver error) would leave the runner thread inside the
+# step while it holds _CARD_LOCK - Stop cannot take effect, the card monitor
+# waits for the lock and a load test hangs in that iteration.  The watchdog
+# watches the runner's progress heartbeat; after _TEST_STEP_TIMEOUT seconds
+# without one it flags the run stuck (error + stop) and tries to cancel the
+# pending transmit by disconnecting the PC/SC link.  That attempt runs in the
+# background because PC/SC clients may serialize the disconnect behind the
+# pending transmit (pcsc-lite locks the context for both calls): when the
+# cancel works, the blocking call raises and the normal step-failure path
+# finalises the run; when it does not, the driver's own timeout eventually
+# returns and the run still finalises as an error, and the status already
+# explains the wait.  The transport is marked stale so the next equip rebuilds
+# it.
+
+_TEST_STEP_TIMEOUT = 30        # seconds without runner progress => stuck card
+_TEST_WATCHDOG = None
+_TEST_WATCHDOG_ABORTED = None  # heartbeat value already aborted
+
+
+def _test_heartbeat():
+    """Mark runner progress for the watchdog.  A bounded STATUS/drain loop may
+    legitimately run longer than one exchange, so it beats once per attempt; a
+    single exchange beats when its step starts and when it finishes (so the
+    gap before the next step or the cleanup drain is covered)."""
+    with _TEST_LOCK:
+        _TEST_RUN['heartbeat'] = time.time()
+
+
+def _abort_card_transport():
+    """Try to cancel a blocking SCardTransmit, then mark the transport stale.
+
+    The disconnect runs on its own daemon thread: PC/SC clients may serialize
+    it behind the pending transmit (pcsc-lite takes the per-context lock in
+    both calls), so calling it here would block the watchdog instead of
+    cancelling anything.  Where the platform does cancel, the blocked call
+    raises and the normal step-failure path finalises the run; where it does
+    not, the run stays flagged stuck until the driver returns.  Never takes
+    _CARD_LOCK - the stuck runner holds it."""
+    global _TRANSPORT_STALE
+    scc = getattr(_server_ref, 'scc', None) if _server_ref else None
+    con = getattr(getattr(scc, '_tp', None), '_con', None)
+    if con is None:
+        sys.stderr.write('TEST-WATCHDOG: no PC/SC connection to abort\n')
+        _TRANSPORT_STALE = True
+        return False
+
+    def _run():
+        global _TRANSPORT_STALE
+        try:
+            con.disconnect()
+        except Exception as e:
+            sys.stderr.write('TEST-WATCHDOG: PC/SC disconnect failed: %s\n' % e)
+        finally:
+            _TRANSPORT_STALE = True
+
+    threading.Thread(target=_run, name='test-abort', daemon=True).start()
+    return True
+
+
+def _test_watchdog_check(now=None):
+    """One watchdog pass.  The heartbeat is the authority (it is beaten at
+    every step start/finish and inside the bounded drains), so a stale one
+    means the runner is stuck inside a card exchange.  Returns the stuck error
+    string when it acted (once per heartbeat), else None.  Split from the
+    loop so the rule is testable."""
+    global _TEST_WATCHDOG_ABORTED
+    now = time.time() if now is None else now
+    with _TEST_LOCK:
+        if not _TEST_RUN.get('running'):
+            _TEST_WATCHDOG_ABORTED = None
+            return None
+        hb = _TEST_RUN.get('heartbeat')
+        if not hb:
+            return None
+        idle = now - hb
+        if idle < _TEST_STEP_TIMEOUT:
+            return None
+        if _TEST_WATCHDOG_ABORTED == hb:
+            return None
+        _TEST_WATCHDOG_ABORTED = hb
+        steps = _TEST_RUN.get('steps') or []
+        cur = steps[-1] if steps else None
+        index = cur.get('index') if cur is not None else _TEST_RUN.get('index')
+        msg = ('the card stopped responding (step %d, no answer for %d s)'
+               % ((index or 0) + 1, int(idle)))
+        _TEST_RUN['stuck'] = True
+        _TEST_RUN['error'] = msg
+        _TEST_RUN['stop'] = True
+    sys.stderr.write('TEST-WATCHDOG: %s - trying to cancel the pending '
+                     'transmit\n' % msg)
+    _abort_card_transport()
+    return msg
+
+
+def _test_watchdog_loop():
+    while True:
+        time.sleep(1.0)
+        try:
+            _test_watchdog_check()
+        except Exception as e:
+            sys.stderr.write('TEST-WATCHDOG error: %s\n' % e)
+
+
+def _test_watchdog_start():
+    """Start the singleton watchdog thread (idempotent; it exits with the
+    process)."""
+    global _TEST_WATCHDOG
+    if _TEST_WATCHDOG is not None and _TEST_WATCHDOG.is_alive():
+        return
+    _TEST_WATCHDOG = threading.Thread(target=_test_watchdog_loop,
+                                      name='test-watchdog', daemon=True)
+    _TEST_WATCHDOG.start()
+
+
 def _test_suite_refs_error(script_store, suite):
    """The cross-store rule: every member script must exist in the script
    store (a suite holds references; the scripts are the single source of
@@ -7457,6 +7632,7 @@ def _test_finish_entry(entry, result):
                'command', 'drained'):
        if result.get(key) is not None:
            fields[key] = result[key]
+    _test_heartbeat()          # the exchange finished: runner progress
    _test_entry_update(entry, fields)


@@ -7550,6 +7726,7 @@ def _test_run_status(scc, step):
    data, sw = '', ''
    used = 0
    for n in range(p['attempts']):
+        _test_heartbeat()
        data, sw = _send_status(scc)
        used = n + 1
        _test_log('STATUS %d/%d -> %s' % (used, p['attempts'], sw))
@@ -7638,6 +7815,7 @@ def _test_run_drain(server, step):
    sw = ''
    polls = 0
    for n in range(p['attempts']):
+        _test_heartbeat()
        data, sw = _send_status(scc)
        polls = n + 1
        _test_log('STATUS %d/%d -> %s' % (polls, p['attempts'], sw))
@@ -8267,6 +8445,7 @@ def _test_drain_pending(scc, fetch_len, result=0x00, max_commands=3):
    answered = []
    try:
        while len(answered) < max_commands:
+            _test_heartbeat()      # a bounded drain still makes progress
            # fetch_len 0 (a 9100 SW) means "length unknown": fetch the full
            # 255 bytes - `%02x` of 0x100 would emit an odd-length APDU
            fdata, sw = scc._tp.send_apdu('%s120000%02x' % (scc.cat_cla,
@@ -8341,6 +8520,7 @@ def _test_run_script(server, script, preset, shared=None):
        entry = _test_entry(step, index)
        with _TEST_LOCK:
            _TEST_RUN['steps'].append(entry)
+            _TEST_RUN['heartbeat'] = time.time()
        # A proactive-drain step is the tool that consumes a pending command
        # (it polls STATUS - the card re-announces the unfetched command -
        # then fetches and answers it), so the guard does not apply to it
@@ -8388,14 +8568,16 @@ def _test_run_script(server, script, preset, shared=None):
                                       'note': str(e)})
            sys.stderr.write('TEST-RUN step %d failed (%s): %s\n'
                             % (index + 1, level, e))
-            if pending is not None:
+            # A stuck card's transport is already being torn down: draining
+            # would only raise again on the dead link.
+            if pending is not None and not _TEST_RUN.get('stuck'):
                with _CARD_LOCK:
                    drained = _test_drain_pending(scc, pending)
                _test_entry_update(entry, {k: v for k, v in drained.items() if k != 'note'})
                pending = None
            if level == 'error':
                break
-    if pending is not None:
+    if pending is not None and not _TEST_RUN.get('stuck'):
        with _CARD_LOCK:
            drained = _test_drain_pending(scc, pending)
        with _TEST_LOCK:
@@ -8423,7 +8605,11 @@ def _test_run_execute(server, script, preset):
        _TEST_RUN['finished'] = time.time()
        if _TEST_RUN['status'] is None:
            levels = [s.get('status') for s in _TEST_RUN['steps']]
-            if stopped:
+            if _TEST_RUN.get('stuck'):
+                # the watchdog's stop flag ended the loop; the run still
+                # failed (a stuck card is not a user stop)
+                _TEST_RUN['status'] = 'error'
+            elif stopped:
                _TEST_RUN['status'] = 'stopped'
            elif 'error' in levels:
                _TEST_RUN['status'] = 'error'
@@ -8462,9 +8648,12 @@ def _test_member_prefix(index, member):
    return '[%d/%d %s %s] ' % (index + 1, total, role, tag)


-def _test_member_status(steps, stopped):
+def _test_member_status(steps, stopped, stuck=False):
    """A suite member's verdict from its step levels (an error stops that
-    script); a stop request marks the member stopped."""
+    script); a stop request marks the member stopped.  A stuck card is an
+    error even though the watchdog's stop flag ended the script."""
+    if stuck:
+        return 'error'
    if stopped:
        return 'stopped'
    levels = [s.get('status') for s in steps]
@@ -8547,6 +8736,7 @@ def _test_run_spawn(server, target, args, thread_name):
    """The shared run start: the run owns the card from here - suspend the
    background poll, answer a paused interactive command, spawn the worker."""
    global _TEST_THREAD, _TEST_RUNNING, _POLL_TIMER
+    _test_watchdog_start()
    _TEST_RUNNING = True
    if _POLL_TIMER is not None:
        _POLL_TIMER.cancel()
@@ -8574,31 +8764,41 @@ def _test_run_spawn(server, target, args, thread_name):
        raise


-def _test_run_start(server, script, preset, script_id=None, adm_info=None):
-    """Start a single-script run."""
+def _test_run_start(server, script, preset, script_id=None, adm_info=None,
+                    iterations=1):
+    """Start a single-script run; `iterations` > 1 makes it a load test."""
    with _TEST_LOCK:
        _TEST_RUN.update({
            'running': True, 'stop': False, 'name': script['name'], 'status': None,
            'session': getattr(server, 'card_session', 0), 'index': 0,
            'total': len(script['steps']), 'steps': [], 'started': time.time(),
-            'finished': None, 'error': None, 'scp80_counter': None,
+            'finished': None, 'error': None, 'stuck': False,
+            'heartbeat': time.time(), 'scp80_counter': None,
            'scp80_counters': {}, 'script_id': script_id,
            'preset': (preset or {}).get('name') or (preset or {}).get('iccid'),
            'suite': None, 'log': [], 'log_prefix': '',
            'adm': adm_info or {'required': False},
+            'load': _test_load_state(iterations),
        })
-    _test_run_spawn(server, _test_run_worker, (server, script, preset), 'test-script')
+    if iterations > 1:
+        sys.stderr.write('TEST-LOAD: script %r, %d iterations\n'
+                         % (script.get('name') or script_id or 'script',
+                            iterations))
+        _test_run_spawn(server, _test_load_script_worker,
+                        (server, script, preset, script_id, iterations), 'test-load')
+    else:
+        _test_run_spawn(server, _test_run_worker, (server, script, preset),
+                        'test-script')


-def _test_suite_execute(server, suite, members, preset, preset_id):
-    """The suite worker body: setup -> members in order (per-member on_fail)
-    -> teardown, one card session, no resume.  The teardown runs on every
-    stop - its purpose is to leave the card ready for a new test - as long as
-    the card is in the reader; a card reset stops the suite and the remaining
-    members are skipped (the report flags it)."""
-    global _TEST_RUNNING
+def _test_suite_execute_once(server, suite, members, preset, preset_id, state,
+                             shared=None):
+    """One suite iteration: setup -> members in order (per-member on_fail) ->
+    teardown, one card session, no resume.  Fills `state` and the live
+    `_TEST_RUN` steps/log; returns ``(status, stopped, session_changed)``.
+    The caller owns the run lifecycle - a load test repeats this per
+    iteration."""
    session_start = getattr(server, 'card_session', 0)
-    state = _TEST_RUN['suite']
    stopped = False
    session_changed = False
    setup_i = next((i for i, (e, _s) in enumerate(members)
@@ -8610,7 +8810,7 @@ def _test_suite_execute(server, suite, members, preset, preset_id):
    # suite, so a second SCP80 script starts from what the first consumed
    # (the store persists each packet's counter; the preset dict here is the
    # suite-start snapshot).
-    shared = {}
+    shared = shared if shared is not None else {}

    def run_member(i):
        entry, script = members[i]
@@ -8620,6 +8820,7 @@ def _test_suite_execute(server, suite, members, preset, preset_id):
            member['started'] = time.time()
            _TEST_RUN['steps'] = []
            _TEST_RUN['index'] = 0
+            _TEST_RUN['heartbeat'] = time.time()
            _TEST_RUN['total'] = len(script['steps'])
            _TEST_RUN['log_prefix'] = _test_member_prefix(i, member)
            log_from = len(_TEST_RUN['log'])
@@ -8628,7 +8829,9 @@ def _test_suite_execute(server, suite, members, preset, preset_id):
            steps = json.loads(json.dumps(_TEST_RUN['steps']))
            log = list(_TEST_RUN['log'][log_from:])
            member.update({
-                'status': _test_member_status(steps, member_stopped),
+                'status': _test_member_status(
+                    steps, member_stopped,
+                    stuck=bool(_TEST_RUN.get('stuck'))),
                'steps': steps,
                'log': log[-_TEST_MEMBER_LOG_MAX:],
                'log_truncated': len(log) > _TEST_MEMBER_LOG_MAX,
@@ -8647,8 +8850,11 @@ def _test_suite_execute(server, suite, members, preset, preset_id):

    abort = False
    if setup_i is not None:
-        status, _s = run_member(setup_i)
-        if status not in ('ok', 'warning'):
+        status, setup_stopped = run_member(setup_i)
+        if setup_stopped:
+            stopped = True
+            abort = True
+        elif status not in ('ok', 'warning'):
            abort = True
            skip_pending('the setup script failed')
    for i in member_idx:
@@ -8664,13 +8870,17 @@ def _test_suite_execute(server, suite, members, preset, preset_id):
            session_changed = True
            skip_pending('card session changed (card removed or re-equipped)')
            break
-        status, _s = run_member(i)
+        status, member_stopped = run_member(i)
+        if member_stopped:
+            stopped = True
+            abort = True
+            break
        if status not in ('ok', 'warning') and members[i][0]['on_fail'] == 'stop':
            abort = True
-    if abort:
-        skip_pending('a previous script failed (on_fail: stop)')
-    else:
+    if stopped:
        skip_pending('stopped')
+    elif abort:
+        skip_pending('a previous script failed (on_fail: stop)')
    if teardown_i is not None:
        if not getattr(server, 'card_present', False):
            with _TEST_LOCK:
@@ -8703,29 +8913,40 @@ def _test_suite_execute(server, suite, members, preset, preset_id):
            'warning': statuses.count('warning'),
            'error': statuses.count('error'),
            'skipped': statuses.count('skipped'),
-            'stopped': bool(stopped),
+            'stopped': bool(stopped or statuses.count('stopped')),
            'session_changed': state['session_changed'],
-            'wall_ms': int((time.time() - (_TEST_RUN['started'] or time.time())) * 1000),
+            'wall_ms': int((time.time() - (state.get('started') or _TEST_RUN['started'] or time.time())) * 1000),
            'counters_before': state.get('counters_before'),
            'counters_after': after,
            'setup': state['members'][setup_i]['status'] if setup_i is not None else None,
            'teardown': state['members'][teardown_i]['status'] if teardown_i is not None else None,
        }
        if 'error' in statuses:
-            _TEST_RUN['status'] = 'error'
+            status = 'error'
        elif 'warning' in statuses:
-            _TEST_RUN['status'] = 'warning'
-        elif stopped:
-            _TEST_RUN['status'] = 'stopped'
+            status = 'warning'
+        elif stopped or 'stopped' in statuses:
+            status = 'stopped'
        else:
-            _TEST_RUN['status'] = 'ok'
+            status = 'ok'
+    return status, stopped, session_changed
+
+
+def _test_suite_execute(server, suite, members, preset, preset_id, shared=None):
+    """The single suite run: one iteration, then the run state is finalised.
+    (A load test calls `_test_suite_execute_once` itself.)"""
+    global _TEST_RUNNING
+    state = _TEST_RUN['suite']
+    status, _stopped, _session_changed = _test_suite_execute_once(
+        server, suite, members, preset, preset_id, state, shared)
+    with _TEST_LOCK:
+        _TEST_RUN['status'] = status
        _TEST_RUN['running'] = False
        _TEST_RUN['finished'] = time.time()
        _TEST_RUN['log_prefix'] = ''      # the report is done: no stale prefix
    _TEST_RUNNING = False
    _sim_resume_poll()

-
 def _test_suite_worker(server, suite, members, preset, preset_id):
    global _TEST_RUNNING
    try:
@@ -8776,6 +8997,7 @@ def _test_suite_state(suite, members, preset, preset_id, session, reset_info=Non
        'require_adm': bool(suite.get('require_adm')),
        'start_reset': bool(suite.get('start_reset')),
        'reset': reset_info or None,
+        'started': time.time(),
        'session_start': session, 'session_end': None,
        'session_changed': False, 'preset_id': preset_id,
        'preset': (preset or {}).get('name') or (preset or {}).get('iccid'),
@@ -8792,7 +9014,7 @@ def _test_suite_state(suite, members, preset, preset_id, session, reset_info=Non


 def _test_suite_start(server, suite, members, preset, preset_id, adm_info=None,
-                      reset_info=None):
+                      reset_info=None, iterations=1):
    """Start a suite run: `members` is the resolved, normalised
    ``[(entry, script)]`` list in run order.  `reset_info` records the
    start-of-suite card reset the caller already performed."""
@@ -8803,17 +9025,214 @@ def _test_suite_start(server, suite, members, preset, preset_id, adm_info=None,
            'name': suite.get('name') or 'test suite', 'status': None,
            'session': session, 'index': 0, 'total': 0, 'steps': [],
            'started': time.time(), 'finished': None, 'error': None,
+            'stuck': False, 'heartbeat': time.time(),
            'scp80_counter': None, 'scp80_counters': {},
            'script_id': None,
            'preset': (preset or {}).get('name') or (preset or {}).get('iccid'),
            'suite': _test_suite_state(suite, members, preset, preset_id, session,
                                       reset_info=reset_info),
            'log': [], 'log_prefix': '', 'adm': adm_info or {'required': False},
+            'load': _test_load_state(iterations, suite=True),
        })
+    if iterations > 1:
+        sys.stderr.write('TEST-LOAD: suite %r, %d iterations\n'
+                         % (suite.get('name') or 'suite', iterations))
+        _test_run_spawn(server, _test_load_suite_worker,
+                        (server, suite, members, preset, preset_id, iterations),
+                        'test-load')
+    else:
        _test_run_spawn(server, _test_suite_worker,
                        (server, suite, members, preset, preset_id), 'test-suite')


+
+def _test_load_finalise(status, done, iterations, failed, attempted=None):
+    """Close a load run: the counters, the timing and the final status.  Only
+    the failing iteration's steps/log stay (successful ones were discarded).
+    `attempted` counts the started iterations (the failing/partial one
+    included), so `avg_ms` is a real per-iteration time."""
+    with _TEST_LOCK:
+        now = time.time()
+        load = _TEST_RUN['load']
+        started = load.get('started') or _TEST_RUN.get('started') or now
+        elapsed = int((now - started) * 1000)
+        attempted = attempted if attempted is not None else done
+        load.update({'done': done, 'failed_iteration': failed, 'finished': now,
+                     'elapsed_ms': elapsed,
+                     'avg_ms': int(elapsed / attempted) if attempted else None})
+        _TEST_RUN['running'] = False
+        _TEST_RUN['finished'] = now
+        _TEST_RUN['status'] = status
+        _TEST_RUN['log_prefix'] = ''
+    if failed:
+        how = 'failed at iteration %d/%d' % (failed, iterations)
+    elif status == 'stopped':
+        how = 'stopped after %d/%d' % (done, iterations)
+    else:
+        how = 'done %d/%d' % (done, iterations)
+    sys.stderr.write('TEST-LOAD: %s (%s)\n'
+                     % (_TEST_RUN.get('name') or 'run', how))
+
+
+def _test_load_script_worker(server, script, preset, script_id, iterations):
+    """Repeat one script `iterations` times: successful runs are discarded,
+    an error (or a stop) keeps its steps/log as the report (v3.37.0)."""
+    global _TEST_RUNNING
+    shared = {}                    # the SCP80 counter continues across runs
+    session = getattr(server, 'card_session', 0)
+    done = 0
+    attempted = 0
+    failed = None
+    final = 'ok'
+    worst = 'ok'
+    try:
+        for i in range(1, iterations + 1):
+            with _TEST_LOCK:
+                if _TEST_RUN['stop']:
+                    final = 'stopped'
+                    break
+                _TEST_RUN.update({'steps': [], 'index': 0, 'status': None,
+                                  'error': None, 'stuck': False,
+                                  'heartbeat': time.time(), 'log': [],
+                                  'log_prefix': ''})
+                _TEST_RUN['load'].update({'iteration': i, 'done': done})
+            attempted = i
+            if getattr(server, 'card_session', 0) != session:
+                with _TEST_LOCK:
+                    _TEST_RUN['error'] = ('the card session changed during the '
+                                          'load test (card removed or re-equipped)')
+                failed = i
+                final = 'error'
+                break
+            _ctx, stopped = _test_run_script(server, script, preset, shared=shared)
+            with _TEST_LOCK:
+                levels = [st.get('status') for st in _TEST_RUN['steps']]
+                if _TEST_RUN.get('stuck'):
+                    status = 'error'
+                elif stopped:
+                    status = 'stopped'
+                elif 'error' in levels:
+                    status = 'error'
+                elif 'warning' in levels:
+                    status = 'warning'
+                else:
+                    status = 'ok'
+                if status in ('error', 'stopped'):
+                    if status == 'error':
+                        failed = i
+                        if not _TEST_RUN.get('error'):
+                            _TEST_RUN['error'] = 'iteration %d failed' % i
+                    final = status
+                else:
+                    if status == 'warning':
+                        worst = 'warning'
+                    done = i
+                    _TEST_RUN['steps'] = []
+                    _TEST_RUN['log'] = []
+                    _TEST_RUN['load']['done'] = done
+            if final != 'ok':
+                break
+        if final == 'ok':
+            final = worst if done == iterations else 'stopped'
+        _test_load_finalise(final, done, iterations, failed, attempted)
+    except Exception as e:
+        # a raised runner error must not leave the run "running" forever (the
+        # PWA would poll a dead run and Clear would stay refused)
+        traceback.print_exc()
+        with _TEST_LOCK:
+            _TEST_RUN['error'] = str(e)
+        if failed is None:
+            failed = _TEST_RUN['load'].get('iteration') or (done + 1)
+        _test_load_finalise('error', done, iterations, failed, attempted)
+    finally:
+        _TEST_RUNNING = False
+        _sim_resume_poll()
+
+
+def _test_load_suite_worker(server, suite, members, preset, preset_id, iterations):
+    """Repeat a whole suite `iterations` times (setup/members/teardown per
+    iteration, one shared counter state); the same success/failure policy as
+    the script load test (v3.37.0)."""
+    global _TEST_RUNNING
+    shared = {}
+    session = getattr(server, 'card_session', 0)
+    done = 0
+    attempted = 0
+    failed = None
+    final = 'ok'
+    worst = 'ok'
+    try:
+        for i in range(1, iterations + 1):
+            with _TEST_LOCK:
+                if _TEST_RUN['stop']:
+                    final = 'stopped'
+                    break
+                # the report's counters_before is this iteration's start; the
+                # protocol state (keys/TARs) stays the run-start snapshot
+                fresh = preset
+                store = getattr(server, 'card_presets', None)
+                if preset_id and store is not None:
+                    fresh = store.get(preset_id) or preset
+                state = _test_suite_state(suite, members, fresh, preset_id, session)
+                _TEST_RUN.update({'suite': state, 'steps': [], 'index': 0,
+                                  'total': 0, 'status': None, 'error': None,
+                                  'stuck': False, 'heartbeat': time.time(),
+                                  'log': [], 'log_prefix': ''})
+                _TEST_RUN['load'].update({'iteration': i, 'done': done})
+            attempted = i
+            if getattr(server, 'card_session', 0) != session:
+                with _TEST_LOCK:
+                    _TEST_RUN['error'] = ('the card session changed during the '
+                                          'load test (card removed or re-equipped)')
+                failed = i
+                final = 'error'
+                break
+            status, stopped, session_changed = _test_suite_execute_once(
+                server, suite, members, preset, preset_id, state, shared)
+            if session_changed:
+                with _TEST_LOCK:
+                    _TEST_RUN['error'] = ('the card session changed during the '
+                                          'load test (card removed or re-equipped)')
+                failed = i
+                final = 'error'
+                break
+            # an error wins over the stop flag: a stuck card sets both, and
+            # the report must say it failed, not that the operator stopped it
+            if status == 'error':
+                failed = i
+                final = 'error'
+                with _TEST_LOCK:
+                    if not _TEST_RUN.get('error'):
+                        _TEST_RUN['error'] = 'iteration %d failed' % i
+                break
+            if stopped or status == 'stopped':
+                final = 'stopped'
+                break
+            if status == 'warning':
+                worst = 'warning'
+            with _TEST_LOCK:
+                done = i
+                _TEST_RUN['load']['done'] = done
+                _TEST_RUN['steps'] = []
+                _TEST_RUN['log'] = []
+                # successful iterations are not tracked (a failure keeps its
+                # own state for the report)
+                _TEST_RUN['suite'] = None
+        if final == 'ok':
+            final = worst if done == iterations else 'stopped'
+        _test_load_finalise(final, done, iterations, failed, attempted)
+    except Exception as e:
+        traceback.print_exc()
+        with _TEST_LOCK:
+            _TEST_RUN['error'] = str(e)
+        if failed is None:
+            failed = _TEST_RUN['load'].get('iteration') or (done + 1)
+        _test_load_finalise('error', done, iterations, failed, attempted)
+    finally:
+        _TEST_RUNNING = False
+        _sim_resume_poll()
+
+
 class PysimHandler(BaseHTTPRequestHandler):
    def _send_json(self, data, status=200):
        self.send_response(status)
@@ -8883,7 +9302,7 @@ class PysimHandler(BaseHTTPRequestHandler):
        self._send_json(resp, 500)
        self._log_resp(resp)

-    def _test_suite_run_request(self, suite_id, body):
+    def _test_suite_run_request(self, suite_id, body, iterations=1):
        """Start a suite run: resolve the members and the preset, run the ADM
        prerequisite, then hand over to the suite runner (one card session,
        no resume)."""
@@ -8994,7 +9413,8 @@ class PysimHandler(BaseHTTPRequestHandler):
        try:
            _test_suite_start(self.server, suite, members, preset,
                              str(body.get('preset_id') or '') or None,
-                              adm_info=adm_info, reset_info=reset_info)
+                              adm_info=adm_info, reset_info=reset_info,
+                              iterations=iterations)
        except Exception as e:
            resp = {'error': 'could not start the suite run: %s' % e}
            self._send_json(resp, 500)
@@ -9618,7 +10038,7 @@ class PysimHandler(BaseHTTPRequestHandler):
                sys.stderr.write('STATUS -> %s\n' % st_sw)
                resp = {'sw': st_sw}
                if st_sw.startswith('91'):
-                    _handle_proactive_chain(scc, st_sw)
+                    _handle_proactive_chain(scc, st_sw, max_rounds=_POLL_CHAIN_MAX)
                    resp['proactive'] = True
                self._send_json(resp)
                self._log_resp(resp)
@@ -11274,9 +11694,16 @@ class PysimHandler(BaseHTTPRequestHandler):
                self._send_json(resp, 409)
                self._log_resp(resp)
                return
+            try:
+                iterations = _test_load_iterations(body)
+            except ValueError as e:
+                resp = {'error': str(e)}
+                self._send_json(resp, 400)
+                self._log_resp(resp)
+                return
            suite_id = str(body.get('suite_id') or '').strip().lower()
            if suite_id:
-                self._test_suite_run_request(suite_id, body)
+                self._test_suite_run_request(suite_id, body, iterations)
                return
            script_raw, script_id, script_err = _test_run_script_from_body(self.server, body)
            if script_err:
@@ -11317,7 +11744,7 @@ class PysimHandler(BaseHTTPRequestHandler):
                return
            try:
                _test_run_start(self.server, script, preset, script_id or None,
-                                adm_info=adm_info)
+                                adm_info=adm_info, iterations=iterations)
            except Exception as e:
                resp = {'error': 'could not start the test run: %s' % e}
                self._send_json(resp, 500)
@@ -11346,6 +11773,7 @@ class PysimHandler(BaseHTTPRequestHandler):
                        'scp80_counters': {},
                        'started': None, 'finished': None, 'error': None,
                        'suite': None, 'log': [], 'log_prefix': '', 'adm': None,
+                        'load': _test_load_state(1),
                    })
            if busy:
                resp = {'error': 'test script is running'}
@@ -146,7 +146,7 @@ class PliDictHttpTests(unittest.TestCase):
        try:
            status, resp = self._req('GET', '/api/pli-dict')
            self.assertEqual(status, 200, resp)
-            self.assertEqual(resp['00'], '130500F1100000')
+            self.assertEqual(resp['00'], '130700F11000010001')
            self.assertEqual(resp['0A'], '630104')
            self.assertEqual(resp['03'], '')      # the host clock answers it
            self.assertEqual(resp['02'], '')      # no default
@@ -1041,7 +1041,10 @@ class TestProactiveDecode(unittest.TestCase):
        # define (TS 102 223 8.x, tags per TS 101 220 V18.3.0 Table 7.23)
        import pysim_simple_server.server as srv
        d = srv.PLI_DEFAULTS
-        self.assertEqual(d[0x00], '130500F1100000')        # 001/01 + LAC 0000
+        # the 7-byte PLI form: 001/01 + LAC 0001 + Cell ID 0001 (length 05
+        # is the Network Rejection event form, TS 31.111 8.19.1)
+        self.assertEqual(d[0x00], '130700F11000010001')
+        self.assertEqual(len(bytes.fromhex(d[0x00])), 9)
        self.assertEqual(d[0x01], '140894104502237315F8')  # IMEI, F pad
        self.assertEqual(d[0x04], '2D02656E')              # 'en'
        self.assertEqual(d[0x05], '2E020000')              # Timing Advance
@@ -1073,6 +1076,19 @@ class TestProactiveDecode(unittest.TestCase):
            srv._PLI_DATA.clear()
            srv._PLI_DATA.update(saved)

+    def test_the_pli_tr_carries_the_7_byte_location(self):
+        # the TR builder injects the location object after the Result; a
+        # 5-byte object (no Cell ID) is the Network Rejection form, not a PLI
+        import pysim_simple_server.server as srv
+        saved = srv._PLI_DATA.pop(0x00, None)
+        try:
+            tr = srv._build_tr(object(), 1, 0x26, 0x81, 0x82, 0x00)
+        finally:
+            if saved is not None:
+                srv._PLI_DATA[0x00] = saved
+        self.assertTrue(tr.startswith(bytes.fromhex('810301260002028281030100')))
+        self.assertTrue(tr.endswith(bytes.fromhex('130700F11000010001')))
+
    def test_decode_cmd_empty_raw(self):
        self.assertEqual(_decode_cmd(0x26, b'', None), [])
        self.assertEqual(_decode_cmd(0x03, None, None), [])
@@ -201,6 +201,20 @@ class TestProactiveChainOwnership(unittest.TestCase):
        S._handle_proactive_chain(scc, '910b', max_rounds=3)
        self.assertEqual(len(self._fetches(scc)), 3)

+    def test_every_chain_call_is_bounded_by_default(self):
+        # v3.37.0: the bound is the default, not only the background poll's
+        # choice - a chatty card must not wedge any call site (the manual
+        # STATUS poll and the ENVELOPE chain used to loop forever)
+        scc = FakeChattyScc()
+        S._handle_proactive_chain(scc, '910b')
+        self.assertEqual(len(self._fetches(scc)), S._POLL_CHAIN_MAX)
+
+    def test_the_manual_status_poll_passes_the_bound(self):
+        import inspect
+        src = inspect.getsource(S.PysimHandler._do_POST)
+        branch = src.split("'/api/status-poll'", 1)[1].split('elif self.path', 1)[0]
+        self.assertIn('max_rounds=_POLL_CHAIN_MAX', branch)
+
    def test_the_chain_stops_when_a_run_owns_the_card(self):
        scc = FakeChattyScc()
        S._TEST_RUNNING = True
@@ -264,12 +278,12 @@ class TestProactiveChainOwnership(unittest.TestCase):
        take.assert_called_once()

    def test_every_run_end_resumes_the_poll(self):
-        # the two worker bodies, their two exception paths and the spawn
-        # failure all call the helper
+        # the two worker bodies, their two exception paths, the spawn failure
+        # and the two load workers (v3.37.0) all call the helper
        import inspect
        src = inspect.getsource(S)
        self.assertEqual(src.count('\n    _sim_resume_poll()'), 2)
-        self.assertEqual(src.count('\n        _sim_resume_poll()'), 3)
+        self.assertEqual(src.count('\n        _sim_resume_poll()'), 5)


 if __name__ == '__main__':
@@ -88,7 +88,8 @@ class BatchStateTests(unittest.TestCase):
                                   'failed_step': 'LOAD (3/7)', 'por_sw': '6A80'})
        srv._ram_batch_record(st, {'iccid': '8970119000004600114', 'result': 'skipped',
                                   'reason': 'no preset matches this ICCID'})
-        self.assertEqual(st['counts'], {'installed': 1, 'failed': 1, 'skipped': 1})
+        self.assertEqual(st['counts'], {'installed': 1, 'already_absent': 0,
+                                       'failed': 1, 'skipped': 1})
        self.assertEqual(st['attempts_total'], 3)
        # the notice names the last attempt's outcome
        self.assertIn('skipped: no preset matches this ICCID', st['notice'])
@@ -230,6 +231,49 @@ class BatchCardTests(unittest.TestCase):
        self.assertEqual(seen['current']['iccid'], '8970119000004600098')
        self.assertIn('installed', st['notice'])           # ...and set for the new outcome

+    def test_a_delete_that_finds_nothing_is_reported_already_absent(self):
+        """A delete run whose DELETE answered 6A88 is a success (the object is
+        gone, the goal is met) but not a deletion: it is counted in its own
+        bucket and the notice says so (v3.36.0)."""
+        st = srv._ram_batch_new_state(_plan('delete'))
+        preset = _preset('8970119000004600098', 'Alfa')
+
+        def fake_run(server, body, plan, scc):
+            return {'success': True, 'final_cntr': '0000000002',
+                    'steps': [{'name': 'DELETE (application)', 'sw': '9000',
+                               'por_status': 'por_ok', 'por_sw': '6A88',
+                               'delete_not_found': True}]}
+
+        with mock.patch.object(srv, '_ram_plan_run', fake_run):
+            srv._ram_batch_handle_card(st, _fake_server('8970119000004600098',
+                                                        _FakeStore([preset])))
+        rec = st['attempts'][0]
+        self.assertEqual(rec['result'], 'already_absent')
+        self.assertTrue(rec['not_found'])
+        self.assertEqual(st['counts']['already_absent'], 1)
+        self.assertEqual(st['counts']['installed'], 0)
+        self.assertIn('already absent', st['notice'])
+        self.assertEqual(st['seen']['8970119000004600098'], 'already_absent')
+
+    def test_a_reinstall_with_a_tolerated_pre_delete_is_still_an_install(self):
+        """The reinstall's pre-delete may tolerate 'not found', but the
+        outcome is the INSTALL, not 'already absent'."""
+        st = srv._ram_batch_new_state(_plan('reinstall-app'))
+        preset = _preset('8970119000004600098', 'Alfa')
+
+        def fake_run(server, body, plan, scc):
+            return {'success': True,
+                    'steps': [{'name': 'DELETE (application)', 'por_sw': '6A88',
+                               'delete_not_found': True},
+                              {'name': 'INSTALL [for install]', 'por_sw': '9000'}]}
+
+        with mock.patch.object(srv, '_ram_plan_run', fake_run):
+            srv._ram_batch_handle_card(st, _fake_server('8970119000004600098',
+                                                        _FakeStore([preset])))
+        self.assertEqual(st['attempts'][0]['result'], 'installed')
+        self.assertEqual(st['counts']['installed'], 1)
+        self.assertEqual(st['counts']['already_absent'], 0)
+
    def test_a_card_already_processed_is_reported_not_reinstalled(self):
        st = srv._ram_batch_new_state(_plan())
        preset = _preset('8970119000004600098', 'Alfa')
@@ -0,0 +1,308 @@
+"""The load-testing run mode (v3.37.0): a script or suite repeated N times.
+
+Successful iterations are discarded (only the count grows); the first error
+keeps that iteration's steps/log as the report; a manual stop or a card session
+change ends the loop.  The workers are driven here with the runner mocked, so
+the loop itself is what the tests pin.
+"""
+
+import copy
+import time
+import unittest
+from unittest import mock
+
+from pysim_simple_server import server as srv
+
+
+class _FakeServer:
+    def __init__(self, session=7):
+        self.card_session = session
+
+
+def _reset_run():
+    srv._TEST_RUN.update({
+        'running': True, 'stop': False, 'name': 'load', 'status': None,
+        'session': 7, 'index': 0, 'total': 0, 'steps': [], 'preset': 'p',
+        'script_id': 's1', 'scp80_counter': None, 'scp80_counters': {},
+        'started': time.time(), 'finished': None, 'error': None,
+        'suite': None, 'log': [], 'log_prefix': '', 'adm': {'required': False},
+        'load': srv._test_load_state(1),
+    })
+
+
+class LoadIterationsTests(unittest.TestCase):
+    def test_default_is_one(self):
+        self.assertEqual(srv._test_load_iterations({}), 1)
+        self.assertEqual(srv._test_load_iterations({'iterations': None}), 1)
+        self.assertEqual(srv._test_load_iterations({'iterations': ''}), 1)
+
+    def test_the_count_is_validated(self):
+        self.assertEqual(srv._test_load_iterations({'iterations': '5'}), 5)
+        self.assertEqual(srv._test_load_iterations({'iterations': 7}), 7)
+        for bad in (0, -1, 1.5, 'abc', srv._TEST_LOAD_MAX + 1, [2]):
+            with self.assertRaises(ValueError):
+                srv._test_load_iterations({'iterations': bad})
+
+    def test_the_state_flags_only_repeated_runs(self):
+        one = srv._test_load_state(1)
+        self.assertFalse(one['enabled'])
+        self.assertIsNone(one['iterations'])
+        many = srv._test_load_state(100)
+        self.assertTrue(many['enabled'])
+        self.assertEqual(many['iterations'], 100)
+        self.assertEqual(many['done'], 0)
+
+
+class ScriptLoadTests(unittest.TestCase):
+    def setUp(self):
+        _reset_run()
+        self._saved = copy.deepcopy(srv._TEST_RUN)
+        srv._TEST_RUNNING = True
+
+    def tearDown(self):
+        srv._TEST_RUN.clear()
+        srv._TEST_RUN.update(self._saved)
+        srv._TEST_RUNNING = False
+
+    def _run(self, statuses, iterations, stop_after=None):
+        """Drive the load worker with `statuses` per iteration; returns the
+        shared dicts passed to the runner and the run counter."""
+        calls = []
+        state = {'i': 0}
+
+        def fake_run(server, script, preset, shared=None):
+            i = state['i']
+            state['i'] += 1
+            calls.append(shared)
+            status = statuses[i] if i < len(statuses) else 'ok'
+            srv._TEST_RUN['steps'].append({'index': 0, 'status': status,
+                                           'label': 'step %d' % (i + 1)})
+            if status == 'error':
+                srv._TEST_RUN['steps'][-1]['note'] = 'boom'
+            if stop_after is not None and i + 1 >= stop_after:
+                srv._TEST_RUN['stop'] = True
+            return {}, status == 'stopped'
+
+        with mock.patch.object(srv, '_test_run_script', fake_run), \
+                mock.patch.object(srv, '_sim_resume_poll'):
+            srv._test_load_script_worker(_FakeServer(),
+                                         {'name': 'S', 'steps': []}, None,
+                                         's1', iterations)
+        return calls
+
+    def test_successful_iterations_are_discarded(self):
+        calls = self._run(['ok'] * 5, 5)
+        load = srv._TEST_RUN['load']
+        self.assertEqual(load['done'], 5)
+        self.assertEqual(load['iteration'], 5)
+        self.assertIsNone(load['failed_iteration'])
+        self.assertEqual(srv._TEST_RUN['status'], 'ok')
+        self.assertEqual(srv._TEST_RUN['steps'], [])      # no tracked results
+        self.assertEqual(srv._TEST_RUN['log'], [])
+        self.assertFalse(srv._TEST_RUN['running'])
+        self.assertIsNotNone(load['elapsed_ms'])
+        self.assertIsNotNone(load['avg_ms'])
+        # the counter state is shared by every iteration
+        self.assertEqual(len(calls), 5)
+        for shared in calls[1:]:
+            self.assertIs(shared, calls[0])
+
+    def test_an_error_stops_the_loop_and_keeps_that_iteration(self):
+        self._run(['ok', 'ok', 'error', 'error', 'error'], 5)
+        load = srv._TEST_RUN['load']
+        self.assertEqual(load['done'], 2)
+        self.assertEqual(load['iteration'], 3)
+        self.assertEqual(load['failed_iteration'], 3)
+        self.assertEqual(srv._TEST_RUN['status'], 'error')
+        self.assertEqual(srv._TEST_RUN['error'], 'iteration 3 failed')
+        self.assertEqual(len(srv._TEST_RUN['steps']), 1)
+        self.assertEqual(srv._TEST_RUN['steps'][0]['note'], 'boom')
+        self.assertFalse(srv._TEST_RUN['running'])
+
+    def test_a_warning_does_not_stop_the_loop(self):
+        self._run(['warning', 'ok', 'warning'], 3)
+        load = srv._TEST_RUN['load']
+        self.assertEqual(load['done'], 3)
+        self.assertIsNone(load['failed_iteration'])
+        self.assertEqual(srv._TEST_RUN['status'], 'warning')
+
+    def test_a_manual_stop_ends_the_loop(self):
+        # the fake sets the stop flag after the second iteration
+        self._run(['ok', 'ok', 'ok'], 50, stop_after=2)
+        load = srv._TEST_RUN['load']
+        self.assertEqual(load['done'], 2)
+        self.assertEqual(srv._TEST_RUN['status'], 'stopped')
+        self.assertFalse(srv._TEST_RUN['running'])
+
+    def test_a_card_session_change_between_iterations_is_an_error(self):
+        calls = []
+        state = {'i': 0}
+
+        def fake_run(server, script, preset, shared=None):
+            state['i'] += 1
+            calls.append(1)
+            srv._TEST_RUN['steps'].append({'index': 0, 'status': 'ok'})
+            server.card_session = 99          # the card was swapped
+            return {}, False
+
+        with mock.patch.object(srv, '_test_run_script', fake_run), \
+                mock.patch.object(srv, '_sim_resume_poll'):
+            srv._test_load_script_worker(_FakeServer(), {'name': 'S', 'steps': []},
+                                         None, 's1', 10)
+        load = srv._TEST_RUN['load']
+        self.assertEqual(len(calls), 1)
+        self.assertEqual(load['failed_iteration'], 2)
+        self.assertEqual(srv._TEST_RUN['status'], 'error')
+        self.assertIn('card session changed', srv._TEST_RUN['error'])
+
+
+class LoadRegressionTests(unittest.TestCase):
+    """The review findings: a raised error, warning aggregation, the average
+    and the per-iteration clearing."""
+
+    def setUp(self):
+        _reset_run()
+        self._saved = copy.deepcopy(srv._TEST_RUN)
+        srv._TEST_RUNNING = True
+
+    def tearDown(self):
+        srv._TEST_RUN.clear()
+        srv._TEST_RUN.update(self._saved)
+        srv._TEST_RUNNING = False
+
+    def test_a_raised_runner_error_finalises_the_run(self):
+        # the PWA would poll a dead run forever and Clear would stay refused
+        def boom(*a, **k):
+            raise OSError('card vanished')
+        with mock.patch.object(srv, '_test_run_script', boom), \
+                mock.patch.object(srv, '_sim_resume_poll'):
+            srv._test_load_script_worker(_FakeServer(), {'name': 'S', 'steps': []},
+                                         None, 's1', 5)
+        self.assertFalse(srv._TEST_RUN['running'])
+        self.assertEqual(srv._TEST_RUN['status'], 'error')
+        self.assertIn('card vanished', srv._TEST_RUN['error'])
+        self.assertEqual(srv._TEST_RUN['load']['failed_iteration'], 1)
+        self.assertIsNotNone(srv._TEST_RUN['load']['finished'])
+
+    def test_a_warning_stays_in_the_final_status(self):
+        statuses = iter(['warning', 'ok', 'ok'])
+        with mock.patch.object(srv, '_test_run_script',
+                               lambda *a, **k: (srv._TEST_RUN['steps'].append(
+                                   {'status': next(statuses)}), ({}, False))[1]), \
+                mock.patch.object(srv, '_sim_resume_poll'):
+            srv._test_load_script_worker(_FakeServer(), {'name': 'S', 'steps': []},
+                                         None, 's1', 3)
+        self.assertEqual(srv._TEST_RUN['status'], 'warning')
+        self.assertEqual(srv._TEST_RUN['load']['done'], 3)
+
+    def test_the_average_covers_the_failing_iteration(self):
+        statuses = iter(['ok', 'ok', 'error'])
+        def fake(*a, **k):
+            st = next(statuses)
+            srv._TEST_RUN['steps'].append({'status': st})
+            return {}, False
+        with mock.patch.object(srv, '_test_run_script', fake), \
+                mock.patch.object(srv, '_sim_resume_poll'):
+            srv._test_load_script_worker(_FakeServer(), {'name': 'S', 'steps': []},
+                                         None, 's1', 10)
+        load = srv._TEST_RUN['load']
+        self.assertEqual(load['done'], 2)
+        self.assertIsNotNone(load['avg_ms'])
+        self.assertLessEqual(load['avg_ms'], load['elapsed_ms'])
+
+    def test_the_steps_are_cleared_before_each_iteration(self):
+        seen = []
+        def fake(*a, **k):
+            seen.append(len(srv._TEST_RUN['steps']))
+            srv._TEST_RUN['steps'].append({'status': 'ok'})
+            return {}, False
+        with mock.patch.object(srv, '_test_run_script', fake), \
+                mock.patch.object(srv, '_sim_resume_poll'):
+            srv._test_load_script_worker(_FakeServer(), {'name': 'S', 'steps': []},
+                                         None, 's1', 4)
+        self.assertEqual(seen, [0, 0, 0, 0], 'every iteration starts empty')
+
+
+class SuiteLoadTests(unittest.TestCase):
+    def setUp(self):
+        _reset_run()
+        self._saved = copy.deepcopy(srv._TEST_RUN)
+        srv._TEST_RUNNING = True
+        self.suite = {'id': 'suite1', 'name': 'S', 'require_adm': False,
+                      'start_reset': False, 'scripts': []}
+        self.members = [({'script_id': 'a', 'role': 'member', 'on_fail': 'stop'},
+                         {'name': 'A', 'steps': []})]
+
+    def tearDown(self):
+        srv._TEST_RUN.clear()
+        srv._TEST_RUN.update(self._saved)
+        srv._TEST_RUNNING = False
+
+    def _run(self, statuses, iterations):
+        calls = []
+        counter = {'i': 0}
+
+        def fake_once(server, suite, members, preset, preset_id, state, shared=None):
+            i = counter['i']
+            counter['i'] += 1
+            calls.append(shared)
+            state['members'][0]['status'] = 'ok'
+            state['summary'] = {'ok': 1, 'warning': 0, 'error': 0, 'skipped': 0,
+                                'stopped': False, 'wall_ms': 1}
+            return statuses[i] if i < len(statuses) else 'ok', False, False
+
+        with mock.patch.object(srv, '_test_suite_execute_once', fake_once), \
+                mock.patch.object(srv, '_sim_resume_poll'):
+            srv._test_load_suite_worker(_FakeServer(), self.suite, self.members,
+                                        None, None, iterations)
+        return calls
+
+    def test_successful_iterations_are_discarded(self):
+        calls = self._run(['ok'] * 4, 4)
+        load = srv._TEST_RUN['load']
+        self.assertEqual(load['done'], 4)
+        self.assertIsNone(srv._TEST_RUN['suite'])          # no tracked reports
+        self.assertEqual(srv._TEST_RUN['steps'], [])
+        self.assertEqual(srv._TEST_RUN['status'], 'ok')
+        self.assertEqual(len(calls), 4)
+        self.assertIs(calls[0], calls[-1])                 # one counter state
+
+    def test_an_error_keeps_the_failing_iteration_report(self):
+        self._run(['ok', 'error', 'error'], 3)
+        load = srv._TEST_RUN['load']
+        self.assertEqual(load['done'], 1)
+        self.assertEqual(load['failed_iteration'], 2)
+        self.assertEqual(srv._TEST_RUN['status'], 'error')
+        self.assertIsNotNone(srv._TEST_RUN['suite'])
+        self.assertEqual(srv._TEST_RUN['suite']['name'], 'S')
+
+    def test_a_session_change_between_suite_iterations_is_an_error(self):
+        def fake_once(server, suite, members, preset, preset_id, state, shared=None):
+            return 'ok', False, True
+        with mock.patch.object(srv, '_test_suite_execute_once', fake_once), \
+                mock.patch.object(srv, '_sim_resume_poll'):
+            srv._test_load_suite_worker(_FakeServer(), self.suite, self.members,
+                                        None, None, 5)
+        self.assertEqual(srv._TEST_RUN['load']['failed_iteration'], 1)
+        self.assertEqual(srv._TEST_RUN['status'], 'error')
+        self.assertIn('card session changed', srv._TEST_RUN['error'])
+        self.assertIsNotNone(srv._TEST_RUN['suite'])
+
+    def test_a_stopped_member_ends_the_suite_loop(self):
+        def fake_once(server, suite, members, preset, preset_id, state, shared=None):
+            return 'ok', True, False
+        with mock.patch.object(srv, '_test_suite_execute_once', fake_once), \
+                mock.patch.object(srv, '_sim_resume_poll'):
+            srv._test_load_suite_worker(_FakeServer(), self.suite, self.members,
+                                        None, None, 5)
+        self.assertEqual(srv._TEST_RUN['status'], 'stopped')
+        self.assertEqual(srv._TEST_RUN['load']['done'], 0)
+
+    def test_a_stopped_iteration_ends_the_loop(self):
+        self._run(['stopped'], 5)
+        self.assertEqual(srv._TEST_RUN['status'], 'stopped')
+        self.assertEqual(srv._TEST_RUN['load']['done'], 0)
+
+
+if __name__ == '__main__':
+    unittest.main()
@@ -0,0 +1,317 @@
+"""The test-runner stuck-card watchdog (v3.37.0).
+
+pyscard's SCardTransmit is blocking and has no timeout: a card that stops
+answering (no SW, no driver error) leaves the runner inside a step holding
+_CARD_LOCK, so Stop cannot take effect and a load test hangs in that iteration.
+The watchdog notices the missing runner progress, fails the run, and
+disconnects the PC/SC link - the pending transmit then raises and the normal
+step-failure path finalises the run.  The transport is marked stale so the next
+equip/operation rebuilds it.
+"""
+
+import copy
+import threading
+import time
+import unittest
+from unittest import mock
+
+from pysim_simple_server import server as srv
+from pysim_simple_server import testscript as T
+
+
+def _warning_script():
+    return T.normalise_script({'name': 'stuck', 'steps': [
+        {'type': 'action', 'kind': 'status', 'params': {'attempts': 1},
+         'check': {'sw': '9000'}, 'on_fail': 'warning'}]},
+        srv._test_command_type)
+
+
+class _RaisingScc:
+    """send_apdu raises like a disconnected PC/SC handle after the abort."""
+    cat_cla = '80'
+
+    def __init__(self):
+        self._tp = self
+
+    def send_apdu(self, apdu):
+        raise RuntimeError('SCARD_E_INVALID_HANDLE')
+
+
+class _FakeServer:
+    def __init__(self, scc=None, session=7):
+        self.scc = scc
+        self.card_session = session
+        self.card_present = True
+        self.sms_oa = '12345'
+        self.sms_sc = '12345678912'
+        self.stk_pending = None
+        self.menu_active = False
+
+
+class _FakeCon:
+    def __init__(self, fail=False):
+        self.calls = 0
+        self.fail = fail
+        self.done = threading.Event()
+
+    def disconnect(self):
+        self.calls += 1
+        try:
+            if self.fail:
+                raise RuntimeError('dead handle')
+        finally:
+            self.done.set()
+
+
+class _FakeScc:
+    def __init__(self, con):
+        self._tp = type('T', (), {'_con': con})()
+
+
+class _WatchdogTestCase(unittest.TestCase):
+    def setUp(self):
+        self.saved_run = copy.deepcopy(srv._TEST_RUN)
+        self.saved_ref = srv._server_ref
+        self.saved_stale = srv._TRANSPORT_STALE
+        self.saved_aborted = srv._TEST_WATCHDOG_ABORTED
+        srv._TEST_WATCHDOG_ABORTED = None
+        srv._TRANSPORT_STALE = False
+
+    def tearDown(self):
+        srv._TEST_RUN.clear()
+        srv._TEST_RUN.update(self.saved_run)
+        srv._server_ref = self.saved_ref
+        srv._TRANSPORT_STALE = self.saved_stale
+        srv._TEST_WATCHDOG_ABORTED = self.saved_aborted
+
+    def _run(self, **fields):
+        srv._TEST_RUN.update({
+            'running': True, 'stop': False, 'status': None, 'error': None,
+            'stuck': False, 'heartbeat': None, 'steps': [], 'index': 0,
+        })
+        srv._TEST_RUN.update(fields)
+
+    @staticmethod
+    def _step(started, index=0, status='running'):
+        return {'index': index, 'status': status, 'started': started}
+
+
+class WatchdogCheckTests(_WatchdogTestCase):
+    def test_an_idle_runner_is_not_watched(self):
+        now = time.time()
+        self._run(running=False, heartbeat=now - 100,
+                  steps=[self._step(now - 100)])
+        with mock.patch.object(srv, '_abort_card_transport') as abort:
+            self.assertIsNone(srv._test_watchdog_check(now))
+        abort.assert_not_called()
+
+    def test_a_step_within_the_timeout_is_ignored(self):
+        now = time.time()
+        started = now - srv._TEST_STEP_TIMEOUT + 5
+        self._run(heartbeat=started, steps=[self._step(started)])
+        with mock.patch.object(srv, '_abort_card_transport') as abort:
+            self.assertIsNone(srv._test_watchdog_check(now))
+        abort.assert_not_called()
+
+    def test_finished_work_with_fresh_progress_is_ignored(self):
+        now = time.time()
+        self._run(heartbeat=now - 1,
+                  steps=[self._step(now - 100, status='ok')])
+        with mock.patch.object(srv, '_abort_card_transport') as abort:
+            self.assertIsNone(srv._test_watchdog_check(now))
+        abort.assert_not_called()
+
+    def test_a_stale_heartbeat_between_steps_still_fires(self):
+        # the bounded drains run with no step entry "running": the heartbeat
+        # is the authority, so a hang there is caught too
+        now = time.time()
+        self._run(heartbeat=now - 100,
+                  steps=[self._step(now - 100, index=4, status='ok')])
+        with mock.patch.object(srv, '_abort_card_transport') as abort:
+            msg = srv._test_watchdog_check(now)
+        self.assertIn('step 5', msg)
+        abort.assert_called_once()
+
+    def test_a_stuck_step_fails_the_run_and_aborts_the_link(self):
+        now = time.time()
+        started = now - srv._TEST_STEP_TIMEOUT - 2
+        self._run(heartbeat=started, steps=[self._step(started)])
+        with mock.patch.object(srv, '_abort_card_transport') as abort:
+            msg = srv._test_watchdog_check(now)
+        self.assertIn('the card stopped responding', msg)
+        self.assertIn('step 1', msg)
+        self.assertIn('no answer for', msg)
+        self.assertTrue(srv._TEST_RUN['stuck'])
+        self.assertTrue(srv._TEST_RUN['stop'])
+        self.assertEqual(srv._TEST_RUN['error'], msg)
+        abort.assert_called_once()
+
+    def test_the_abort_happens_once_per_heartbeat(self):
+        now = time.time()
+        started = now - 100
+        self._run(heartbeat=started, steps=[self._step(started)])
+        with mock.patch.object(srv, '_abort_card_transport') as abort:
+            self.assertIsNotNone(srv._test_watchdog_check(now))
+            self.assertIsNone(srv._test_watchdog_check(now + 1))
+            self.assertIsNone(srv._test_watchdog_check(now + 2))
+        abort.assert_called_once()
+
+    def test_new_progress_rearms_the_watchdog(self):
+        now = time.time()
+        started = now - 100
+        self._run(heartbeat=started, steps=[self._step(started)])
+        with mock.patch.object(srv, '_abort_card_transport') as abort:
+            self.assertIsNotNone(srv._test_watchdog_check(now))
+            # the runner made progress (a new beat) and then hung again
+            srv._TEST_RUN['steps'].append(self._step(started + 1, index=1))
+            srv._TEST_RUN['heartbeat'] = started + 1
+            self.assertIsNotNone(srv._test_watchdog_check(now))
+        self.assertEqual(abort.call_count, 2)
+
+    def test_a_heartbeat_keeps_a_long_poll_loop_alive(self):
+        # a bounded STATUS/drain loop may legitimately run longer than one
+        # exchange: its per-attempt heartbeat must keep the watchdog quiet
+        now = time.time()
+        self._run(heartbeat=now - 1, steps=[self._step(now - 120)])
+        with mock.patch.object(srv, '_abort_card_transport') as abort:
+            self.assertIsNone(srv._test_watchdog_check(now))
+        abort.assert_not_called()
+
+
+class TransportAbortTests(_WatchdogTestCase):
+    def _wait_stale(self):
+        for _ in range(200):
+            if srv._TRANSPORT_STALE:
+                return True
+            time.sleep(0.01)
+        return False
+
+    def test_the_connection_is_disconnected_and_marked_stale(self):
+        # the disconnect is attempted on a background thread (a synchronous
+        # one would block behind the pending transmit on pcsc-lite)
+        con = _FakeCon()
+        srv._server_ref = _FakeServer(_FakeScc(con))
+        self.assertTrue(srv._abort_card_transport())
+        self.assertTrue(con.done.wait(2))
+        self.assertEqual(con.calls, 1)
+        self.assertTrue(self._wait_stale())
+
+    def test_a_missing_connection_is_not_fatal(self):
+        srv._server_ref = _FakeServer(None)
+        self.assertFalse(srv._abort_card_transport())
+        self.assertTrue(srv._TRANSPORT_STALE)
+
+    def test_a_failing_disconnect_is_swallowed(self):
+        con = _FakeCon(fail=True)
+        srv._server_ref = _FakeServer(_FakeScc(con))
+        self.assertTrue(srv._abort_card_transport())
+        self.assertTrue(con.done.wait(2))
+        self.assertEqual(con.calls, 1)
+        self.assertTrue(self._wait_stale())
+
+
+class HeartbeatTests(_WatchdogTestCase):
+    def test_a_finished_step_beats(self):
+        # between steps the heartbeat must stay fresh (the drains run with no
+        # step entry "running")
+        self._run(heartbeat=0)
+        entry = {'index': 0, 'started': time.time() - 5}
+        before = time.time()
+        srv._test_finish_entry(entry, {'status': 'ok'})
+        self.assertGreaterEqual(srv._TEST_RUN['heartbeat'], before)
+
+
+class StuckStatusTests(_WatchdogTestCase):
+    def test_a_stuck_member_is_an_error_not_a_stop(self):
+        self.assertEqual(
+            srv._test_member_status([{'status': 'warning'}], True, stuck=True),
+            'error')
+        self.assertEqual(
+            srv._test_member_status([{'status': 'warning'}], True),
+            'stopped')
+
+    def test_a_stuck_run_finalises_as_error(self):
+        script = _warning_script()
+        self._run(stuck=True, heartbeat=time.time(), stop=True,
+                  total=1, name='stuck')
+        srv._test_run_execute(_FakeServer(_RaisingScc()), script, {})
+        self.assertEqual(srv._TEST_RUN['status'], 'error')
+        self.assertFalse(srv._TEST_RUN['running'])
+
+    def test_the_control_warning_run_without_the_stuck_flag(self):
+        # the same raising warning step is only a warning when the watchdog
+        # did not mark the run: the stuck flag is what changes the verdict
+        script = _warning_script()
+        self._run(stuck=False, heartbeat=time.time(), total=1, name='stuck')
+        srv._test_run_execute(_FakeServer(_RaisingScc()), script, {})
+        self.assertEqual(srv._TEST_RUN['status'], 'warning')
+
+    def test_a_blocked_runner_step_ends_the_run_as_error(self):
+        # end to end: a step blocked inside send_apdu is flagged by the
+        # watchdog; when the call returns (the driver's timeout / the abort)
+        # the ordinary failure path finalises the run as an error even though
+        # the step was a warning one
+        entered = threading.Event()
+        release = threading.Event()
+
+        class _BlockingScc:
+            cat_cla = '80'
+
+            def __init__(self):
+                self._tp = self
+
+            def send_apdu(self, apdu):
+                entered.set()
+                release.wait(5)
+                raise RuntimeError('SCARD_E_NOT_TRANSACTED')
+
+        script = _warning_script()
+        self._run(total=1, name='stuck')
+        saved_timeout = srv._TEST_STEP_TIMEOUT
+        srv._TEST_STEP_TIMEOUT = 0.2
+        try:
+            runner = threading.Thread(
+                target=srv._test_run_execute,
+                args=(_FakeServer(_BlockingScc()), script, {}))
+            runner.start()
+            self.assertTrue(entered.wait(2))
+            time.sleep(0.3)
+            with mock.patch.object(srv, '_abort_card_transport') as abort:
+                msg = srv._test_watchdog_check()
+            self.assertIn('stopped responding', msg)
+            abort.assert_called_once()
+            release.set()
+            runner.join(5)
+        finally:
+            srv._TEST_STEP_TIMEOUT = saved_timeout
+        self.assertFalse(runner.is_alive())
+        self.assertTrue(srv._TEST_RUN['stuck'])
+        self.assertEqual(srv._TEST_RUN['status'], 'error')
+        self.assertFalse(srv._TEST_RUN['running'])
+
+    def test_a_stuck_load_iteration_is_a_failure(self):
+        def fake_run(server, script, preset, shared=None):
+            srv._TEST_RUN['steps'].append(
+                {'index': 0, 'status': 'warning', 'label': 's'})
+            srv._TEST_RUN['stuck'] = True
+            srv._TEST_RUN['error'] = ('the card stopped responding '
+                                      '(step 1, no answer for 31 s)')
+            return {}, True
+
+        self._run(load=srv._test_load_state(3), name='load')
+        srv._TEST_RUN['load'].update({'iteration': 1})
+        with mock.patch.object(srv, '_test_run_script', fake_run), \
+                mock.patch.object(srv, '_sim_resume_poll'):
+            srv._test_load_script_worker(_FakeServer(), {'name': 'S',
+                                                         'steps': []}, None,
+                                         's1', 3)
+        load = srv._TEST_RUN['load']
+        self.assertEqual(srv._TEST_RUN['status'], 'error')
+        self.assertEqual(load['failed_iteration'], 1)
+        self.assertEqual(load['done'], 0)
+        self.assertIn('stopped responding', srv._TEST_RUN['error'])
+        self.assertFalse(srv._TEST_RUN['running'])
+
+
+if __name__ == '__main__':
+    unittest.main()
Author	SHA1	Message	Date
catarrh	513f16d9d2	PLI: return the 7-byte location object, bound every proactive chain release / release (push) Waiting to run Details The PLI answer used the 5-byte form 13 05 <MCC/MNC> <LAC> with LAC 0000 and no Cell ID. That is the Network Rejection event download form (TS 31.111 8.19.1); a PLI location for GERAN is 7 bytes - MCC/MNC, LAC and Cell ID. A card applet that re-reads its location (the Alfa MultiIMSI applet, after a *LOCI status change) treated it as empty and re-requested PLI on every STATUS tick. The default is now 13 07 00F110 0001 0001: PLMN 001/01 (the test network), LAC 0001, Cell ID 0001. _build_tr injects it after the Result; the TR data is pinned by a test. While verifying a live PLI, the manual /api/status-poll chain turned out to be unbounded: it looped on the PLI storm while holding _CARD_LOCK, so the request never returned and every card operation waited. _handle_proactive_chain's max_rounds now defaults to _POLL_CHAIN_MAX (the background poll always passed it; the manual poll, ENVELOPE, menu and TP chains did not), and the manual poll passes it explicitly. A card that re-queues a command can no longer wedge the card lock from any call site. Tests: the PLI default + the built TR, the default chain bound and the manual-poll call pin. 1061 Python + 1069 frontend; sw cache simple-v477.	2026-10-24 01:51:35 +03:00
catarrh	643c6609f3	test runner: add the stuck-card watchdog A card that stops answering (no SW, no driver error) leaves the runner inside a blocking SCardTransmit while it holds the card lock - Stop cannot take effect, the presence monitor waits for the lock and a load test hangs in that iteration without explanation. The runner now beats a progress heartbeat (each step start/finish and each attempt of a bounded STATUS/drain loop); after _TEST_STEP_TIMEOUT (30 s) the watchdog flags the run `stuck`, records the step and the idle time, sets the stop flag and tries to cancel the blocked transmit by disconnecting the PC/SC link on a daemon thread. The disconnect must run in the background: pcsc-lite serializes it behind the pending transmit (both take the per-context lock), so a synchronous call would block the watchdog instead of cancelling anything. Where a platform does cancel, the step fails through the ordinary on_fail path; either way the finalisers map `stuck` to `error` - a stuck card is not a user stop - and the pending drain is skipped on the dead link. The transport is marked stale once the disconnect returns, so the next equip rebuilds it. The run status gains `stuck` and `heartbeat`. Help EN/RU, docs/test-scripts.md, docs/api.md, the CHANGELOG (folded into the unreleased 3.37.0) and both AGENTS describe the behaviour and the background-disconnect caveat; the sw cache is simple-v476 (help is precached). Tests: tests/test_test_watchdog.py (17: the check rule, the abort, the final status of the single/load/suite runs and a blocked-runner integration case). 1058 Python + 1069 frontend.	2026-10-24 00:43:39 +03:00
catarrh	2181ef6bbf	v3.37.0: load testing for test scripts and suites POST /api/test/run takes an optional iterations count (the PWA's Iterations field): the run repeats, successful iterations are discarded (steps/log cleared, only load.done and the timing grow), the first error stops the loop and keeps that iteration's report, a stop ends it after the current iteration, and a card session change between iterations is an error. The SCP80 counter state is shared across iterations, so packets continue from what the previous run consumed. The status gains a load block (iteration/done/failed_iteration/ elapsed_ms/avg_ms; disabled for a normal run). The suite runner was split into a per-iteration body (_test_suite_execute_once) + the single-run finaliser, and two load workers own the loops. Review fixes: a raised runner error finalises as error (no stuck running state), a suite-internal session change is an error, a stopped member ends the suite loop, the final status keeps a warning from any iteration, avg_ms covers the attempted iterations, and each suite iteration's counters_before is read fresh. 1041 Python + 1069 frontend; simple-v475.	2026-10-23 23:44:42 +03:00
catarrh	851ff65af6	changelog: note the README Telegram channel link release / release (push) Waiting to run Details	2026-10-23 21:46:10 +03:00
catarrh	68a22e854b	readme: link the project news Telegram channel Add @youeyeseesee (releases and updates for my GitHub projects) near the top of both READMEs.	2026-10-23 21:44:45 +03:00
catarrh	39d384636e	batch waiting cue: move the contact pad to the bottom edge The card's contact pad sat mid-card; it now sits near the bottom edge, away from the top-right cut corner (the cut side is the card's top). No other geometry or animation changes. 1024 Python + 1063 frontend; simple-v473.	2026-10-23 21:41:30 +03:00
catarrh	14d15b1fe8	batch waiting cue: use the app's SIM silhouette for the card Redraw the inline card-into-reader SVG after sim.svg: a rounded card with the top-right cut corner and the 3x2 contact pad, sliding behind the reader bar. The viewBox grew to 72x48 and the resting position is 0, so the card is no longer clipped at the top; the insert slide is 20px. The currentColor/Tailwind theming and the reduced-motion fallback are unchanged. 1024 Python + 1063 frontend; simple-v472.	2026-10-23 21:37:15 +03:00
catarrh	110dc1f1ad	batch delete: report an already-absent object as such A delete run whose DELETE answered 6A88/6A82 is still a success (the object is gone, the goal is met) but it is no longer reported as a plain "OK": the attempt result is 'already_absent', the counts gain that bucket ("Deleted: N - Already absent: M"), the verdict reads "Already absent" and the notice says the card was already absent. A reinstall's tolerated pre-delete keeps the install outcome (the install is the point). 1024 Python + 1063 frontend; simple-v471.	2026-10-23 21:23:55 +03:00
catarrh	fb65644328	v3.36.0: a deliberate "insert the next card" cue for batch runs The waiting phase is the only moment the operator must act, so it now: - shows a card-into-reader SVG animation in the status modal (inline SVG, CSS keyframes in src/style.css; reduced-motion users get the static card-above-the-slot picture); - plays a short Web Audio chime once per card, edge-triggered in the 2 s poll so it never repeats (the AudioContext is unlocked by the Start click); - sets the tab title to "Insert next card" while the window is hidden; - offers a speaker toggle next to the status (persisted as simple_ram_batch_sound, on by default). The prebuilt style.css was rebuilt: it had gone stale and lacked utilities the UI already used (bg-blue-50 on the batch notice, inline-flex, mx-2, max-h-48, w-44, min-w-24). 1022 Python + 1062 frontend green; simple-v470.	2026-10-23 21:14:32 +03:00