From 66cc6f07aef7e287b8f58c36dc42af6625d68f39 Mon Sep 17 00:00:00 2001
From: Adam <adam.gaborjani.szabo@gmail.com>
Date: Sun, 1 Nov 2026 12:12:15 +0000
Subject: [PATCH] Triage the 2026-11-01 bug bash: name sides, reject blank
 input, stop false-red explore jobs

- Current Debate rows show the side (a side column leaves out its own), so a
  Pro and a Con case saved the same minute no longer read as one case listed
  twice. Load buttons name the kind and side as well as the motion.
- Generate, Brainstorm and Attack treat a motion or argument of only spaces as
  blank instead of spending an AI request on it.
- The Sparring progress bar's accessible name fills in the time instead of
  reading "{{time}} remaining".
- The bug-bash Exploration step passes with a warning when e2e explore exits 1
  with issues recorded, and fails on any other error, in place of
  continue-on-error, which marked every job red and also hid real harness
  failures.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
---
 .../features/current-debate.md                |  2 +-
 .../recipes/current-debate.spec.ts            |  2 ++
 .github/workflows/agentic-e2e.yml             | 25 +++++++++++++------
 docs/testing.md                               |  2 +-
 e2e/full-stack/current-debate.spec.ts         |  2 ++
 e2e/quality-backlog.spec.ts                   | 19 ++++++++++++++
 packages/core/src/ai/schemas/attack.ts        |  2 +-
 packages/core/src/ai/schemas/brainstorm.ts    |  2 +-
 packages/core/src/ai/schemas/generate.ts      |  2 +-
 .../features/sparring/SparringRound.tsx       |  2 +-
 src/pages/CurrentDebatePage.tsx               | 13 +++++++---
 11 files changed, 56 insertions(+), 17 deletions(-)

diff --git a/.claude/skills/verify-debatelab/features/current-debate.md b/.claude/skills/verify-debatelab/features/current-debate.md
index fa7e63d70..495af1908 100644
--- a/.claude/skills/verify-debatelab/features/current-debate.md
+++ b/.claude/skills/verify-debatelab/features/current-debate.md
@@ -30,7 +30,7 @@ Preconditions:
 - **Scope.** Add a source on `/app/library` with a unique marker (see [library-sources](./library-sources.md)); the test id `library-debate` names the motion. Run Brainstorm or Generate: the motion field already holds the debate's motion and no dialog opens. End state: the newest `providerCalls()` entry for the feature has this debate's marker in `system` and not another debate's. `sources.debate_id`, `cases.debate_id`, and `brainstorms.debate_id` point at the debate they were made in.
 - **Offer.** With no debate, fill a tool's motion and submit. The dialog `Start a debate with this motion?` contains the motion. Click its `Start debate`. End state: the result shows, the strip reads `Working in: <motion>`, and `debates` holds exactly that motion.
 - **Use once.** With a debate, submit a different motion. The dialog `This is a different motion` opens. `chooseForMotion(page, "once")`. End state: the provider call's `system` has no debate marker, and the header still names the old debate.
-- **Dashboard.** `/app/debate`: `current-debate-motion` has the motion, `debate-side-pro` reads `In favor`, `debate-side-con` reads `Against`, and `debate-sources` and `debate-recent` list only this debate's work.
+- **Dashboard.** `/app/debate`: `current-debate-motion` has the motion, `debate-side-pro` reads `In favor`, `debate-side-con` reads `Against`, and `debate-sources` and `debate-recent` list only this debate's work. Each row reads `<kind> · <side> · <time>` (a side column leaves out its own side), and its Load button's name ends with the title, kind and side.
 - **Switch.** Click `debateControl(page)`, then the dialog's `Switch debate`, then `Switch to <motion>`. End state: the Library shows only that debate's sources. In History, the `Brainstorms` and `Cases` tabs show only its results.
 - **Tour.** Seed `debate-trainer-onboarding` with `hasCompletedOnboarding: false` once per session (see the recipe), sign in, and click the tooltip's `Start debate`. Fill `Motion`, click `Start debate`, click `[data-test-id="button-primary"]` twice, then `Get started`. End state: one `debates` row is current, and Generate opens prefilled with no dialog.
 
diff --git a/.claude/skills/verify-debatelab/recipes/current-debate.spec.ts b/.claude/skills/verify-debatelab/recipes/current-debate.spec.ts
index dcd4803b2..b3e7cfcf1 100644
--- a/.claude/skills/verify-debatelab/recipes/current-debate.spec.ts
+++ b/.claude/skills/verify-debatelab/recipes/current-debate.spec.ts
@@ -118,6 +118,8 @@ test("debate-start, debate-scope, debate-once, debate-switch: two debates keep t
   await expect(page.getByTestId("debate-sources")).toContainText("Zoo study");
   await expect(page.getByTestId("debate-sources")).not.toContainText("Homework study");
   await expect(page.getByTestId("debate-recent")).toContainText(ZOOS);
+  // Every row is titled with the motion, so the side tells a Pro case from a Con one.
+  await expect(page.getByTestId("debate-recent")).toContainText("Case · In favor");
   await evidence.shot("05-zoo-dashboard");
 
   // Switching back switches the Library and History with it.
diff --git a/.github/workflows/agentic-e2e.yml b/.github/workflows/agentic-e2e.yml
index 4deeca79b..5e2cded1d 100644
--- a/.github/workflows/agentic-e2e.yml
+++ b/.github/workflows/agentic-e2e.yml
@@ -152,9 +152,9 @@ jobs:
     needs: plan-explore
     runs-on: ubuntu-latest
     timeout-minutes: 45
-    # Findings are leads to reproduce, not verdicts: a red exploration does not
-    # fail the workflow. docs/testing.md has the bug-bash procedure.
-    continue-on-error: true
+    # Findings are leads to reproduce, not verdicts, so they pass with a warning;
+    # a broken run (crash, missing key, stack down) fails the job.
+    # docs/testing.md has the bug-bash procedure.
     strategy:
       fail-fast: false
       matrix:
@@ -198,10 +198,21 @@ jobs:
           GOAL: ${{ matrix.leg.goal }}
           PERSONA: ${{ matrix.leg.persona }}
           TARGET: ${{ matrix.leg.target }}
-        run: >-
-          .claude/skills/verify-debatelab/bin/verify.sh e2e explore "$GOAL"
-          --session student --agent "$PERSONA" --target "$TARGET"
-          --trace=on --video=on --reporter list,markdown
+        # e2e explore exits 1 when it reports an issue, and also when no step ran.
+        # Only the first is the bug bash's output; every other non-zero exit is
+        # the harness or the stack failing.
+        run: |
+          status=0
+          .claude/skills/verify-debatelab/bin/verify.sh e2e explore "$GOAL" \
+            --session student --agent "$PERSONA" --target "$TARGET" \
+            --trace=on --video=on --reporter list,markdown || status=$?
+          report=$(ls -d .verify/artifacts/*-e2e-explore-*/report.json 2>/dev/null | tail -n 1)
+          issues=$(jq '[.run.explore.findings[]? | select(.kind == "issue")] | length' "$report" 2>/dev/null || echo 0)
+          if [ "$status" -eq 1 ] && [ "${issues:-0}" -gt 0 ]; then
+            echo "::warning title=Exploration findings::$PERSONA on $TARGET reported $issues issues to reproduce; see the job summary."
+            exit 0
+          fi
+          exit "$status"
 
       - name: Publish the report
         if: always()
diff --git a/docs/testing.md b/docs/testing.md
index 367625a90..c82958618 100644
--- a/docs/testing.md
+++ b/docs/testing.md
@@ -269,7 +269,7 @@ To write a test, import `test`, `expect`, and the fixtures from [`agentic/suppor
 
 A bug bash runs several explorations in parallel, one charter per area and persona, then proves each finding with a repro test. Ask an agent to "bug bash <area>"; the `e2e` skill's `bug-bash` topic has the procedure. A finding counts only when its repro test fails for the reported reason. Tag repro tests `{ tags: ["repro"] }` so the scheduled suite skips them until the fix lands, then remove the tag so they become regression tests. The client runs on a Vite dev server, so a route that is still compiling can look like a dead link to an explorer. Reproduce before you believe it.
 
-[`.github/workflows/agentic-e2e.yml`](../.github/workflows/agentic-e2e.yml) runs the suite on Wednesdays at 03:47 UTC and on demand (`gh workflow run agentic-e2e.yml --ref <branch>`), never on pull requests, in four jobs: the page sweep and everything else, on each viewport. On the 1st of each month at 05:13 UTC it runs a bug bash instead: one exploration per persona, each with its own charter. On demand with a goal, it runs one exploration with the chosen persona and viewport. It uses the `LIVE_OPENROUTER_API_KEY` secret (the `E2E_AGENT_MODEL` and `E2E_JUDGE_MODEL` repository variables override the models), fails when the key is missing, and writes the report to the job summary. Explorations never fail the workflow: their findings are leads to reproduce. The suite is not a required check: model judgments can flake, and a red weekly run is a finding to triage.
+[`.github/workflows/agentic-e2e.yml`](../.github/workflows/agentic-e2e.yml) runs the suite on Wednesdays at 03:47 UTC and on demand (`gh workflow run agentic-e2e.yml --ref <branch>`), never on pull requests, in four jobs: the page sweep and everything else, on each viewport. On the 1st of each month at 05:13 UTC it runs a bug bash instead: one exploration per persona, each with its own charter. On demand with a goal, it runs one exploration with the chosen persona and viewport. It uses the `LIVE_OPENROUTER_API_KEY` secret (the `E2E_AGENT_MODEL` and `E2E_JUDGE_MODEL` repository variables override the models), fails when the key is missing, and writes the report to the job summary. Findings never fail the workflow: an exploration that reports issues (`e2e explore` exit 1) passes with a warning annotation, because its findings are leads to reproduce. Any other non-zero exit, or exit 1 with no issue recorded, is the harness or the stack failing, and fails the job. The suite is not a required check: model judgments can flake, and a red weekly run is a finding to triage.
 
 e2e is pre-1.0 and its minor releases have broken APIs, so it is pinned exactly. Read its changelog before upgrading, refresh the vendored skill as its header says, and expect the replay cache to re-record after a minor bump.
 
diff --git a/e2e/full-stack/current-debate.spec.ts b/e2e/full-stack/current-debate.spec.ts
index a3defe3b2..ddcc249d0 100644
--- a/e2e/full-stack/current-debate.spec.ts
+++ b/e2e/full-stack/current-debate.spec.ts
@@ -97,6 +97,8 @@ test("two debates keep their motions, sources, and results apart", async ({ sign
   await expect(page.getByTestId("debate-sources")).toContainText("Zoo study");
   await expect(page.getByTestId("debate-sources")).not.toContainText("Homework study");
   await expect(page.getByTestId("debate-recent")).toContainText(ZOOS);
+  // Every row is titled with the motion, so the side tells a Pro case from a Con one.
+  await expect(page.getByTestId("debate-recent")).toContainText("Case · In favor");
   await snap(page, "4-zoo-dashboard");
 
   // Switching back switches the Library and History with it.
diff --git a/e2e/quality-backlog.spec.ts b/e2e/quality-backlog.spec.ts
index 9b126584c..3926c9ad8 100644
--- a/e2e/quality-backlog.spec.ts
+++ b/e2e/quality-backlog.spec.ts
@@ -159,3 +159,22 @@ test('language menu closes once a language is chosen', async ({ page }) => {
   await expect(page.getByRole('radiogroup')).toHaveCount(0);
   await expect(page.locator('html')).toHaveAttribute('lang', 'hu');
 });
+
+// Bug bash 2026-11-01: a motion or argument of only spaces passed validation and spent an AI request.
+test('a motion or argument of only spaces counts as blank', async ({ page }) => {
+  await stubSupabaseSession(page);
+  const aiCalls: string[] = [];
+  page.on('request', (request) => { if (request.url().includes('/functions/v1/ai')) aiCalls.push(request.url()); });
+
+  await page.goto('/app/generate');
+  await page.getByLabel('Motion / topic *').fill('     ');
+  await page.getByRole('button', { name: 'Generate debate case' }).click();
+  await expect(page.getByText('Motion is required')).toBeVisible();
+
+  await page.goto('/app/attack');
+  await page.getByRole('textbox', { name: /argument/i }).first().fill('   ');
+  await page.getByRole('button', { name: 'Find weaknesses' }).click();
+  await expect(page.getByText('Case text is required')).toBeVisible();
+
+  expect(aiCalls).toEqual([]);
+});
diff --git a/packages/core/src/ai/schemas/attack.ts b/packages/core/src/ai/schemas/attack.ts
index 198a19543..ddc1ad85c 100644
--- a/packages/core/src/ai/schemas/attack.ts
+++ b/packages/core/src/ai/schemas/attack.ts
@@ -3,7 +3,7 @@ import { z } from 'zod';
 
 export const AttackInputSchema = z.object({
   // A catalog key, not text: the form shows it in the interface language.
-  caseText: z.string().min(1, 'validation.caseTextRequired'),
+  caseText: z.string().trim().min(1, 'validation.caseTextRequired'),
   aggressiveness: z.enum(['mild', 'moderate', 'ruthless']),
   judgeParadigm: z.enum(['logic', 'values', 'policy']),
   detectFallacies: z.boolean(),
diff --git a/packages/core/src/ai/schemas/brainstorm.ts b/packages/core/src/ai/schemas/brainstorm.ts
index 1de065c10..f289d53e9 100644
--- a/packages/core/src/ai/schemas/brainstorm.ts
+++ b/packages/core/src/ai/schemas/brainstorm.ts
@@ -3,7 +3,7 @@ import { z } from 'zod';
 
 export const BrainstormInputSchema = z.object({
   // A catalog key, not text: the form shows it in the interface language.
-  motion: z.string().min(1, 'validation.motionRequired'),
+  motion: z.string().trim().min(1, 'validation.motionRequired'),
   side: z.enum(['pro', 'con', 'both']),
   context: z.string().optional(),
   outputMode: z.enum(['mindmap', 'buckets', 'topfive']),
diff --git a/packages/core/src/ai/schemas/generate.ts b/packages/core/src/ai/schemas/generate.ts
index 10f14c73b..b4dfa7e2b 100644
--- a/packages/core/src/ai/schemas/generate.ts
+++ b/packages/core/src/ai/schemas/generate.ts
@@ -3,7 +3,7 @@ import { z } from 'zod';
 
 export const GenerateInputSchema = z.object({
   // A catalog key, not text: the form shows it in the interface language.
-  motion: z.string().min(1, 'validation.motionRequired'),
+  motion: z.string().trim().min(1, 'validation.motionRequired'),
   side: z.enum(['pro', 'con']),
   format: z.enum(['custom', 'bp', 'policy', 'classroom']), // Kept others for backward compatibility but UI will restrict
   role: z.enum(['intro', 'r1', 'r2', 'conc', 'general']),
diff --git a/src/components/features/sparring/SparringRound.tsx b/src/components/features/sparring/SparringRound.tsx
index 2856a8ef9..ac0be85bb 100644
--- a/src/components/features/sparring/SparringRound.tsx
+++ b/src/components/features/sparring/SparringRound.tsx
@@ -204,7 +204,7 @@ export function SparringRound({
                   {!judging && <ClockAnnouncer secondsLeft={session.timerRemainingSeconds} running={session.timerRunning} />}
                 </div>
               </div>
-              <Progress value={judging ? 0 : timerProgress} className={cn('mt-5 h-2', tone === 'over' ? '[&>div]:bg-destructive' : isUrgent ? '[&>div]:bg-amber-500' : '[&>div]:bg-primary')} aria-label={t('sparring.round.timeRemaining', 'Time remaining')} />
+              <Progress value={judging ? 0 : timerProgress} className={cn('mt-5 h-2', tone === 'over' ? '[&>div]:bg-destructive' : isUrgent ? '[&>div]:bg-amber-500' : '[&>div]:bg-primary')} aria-label={t('sparring.round.timeRemaining', '{{time}} remaining', { time: formatRoundTime(session.timerRemainingSeconds) })} />
               {awaitingStart && (
                 <div className="mt-4 flex flex-wrap items-center gap-3" data-testid="sparring-ready">
                   {!voice && (
diff --git a/src/pages/CurrentDebatePage.tsx b/src/pages/CurrentDebatePage.tsx
index 108c3c20a..43885ab89 100644
--- a/src/pages/CurrentDebatePage.tsx
+++ b/src/pages/CurrentDebatePage.tsx
@@ -28,22 +28,27 @@ function ViewAll({ to, count, shown }: { to: string; count: number; shown: numbe
   );
 }
 
-function ItemRow({ item }: { item: DebateItem }) {
+// Every row's title is the motion, so the side is what tells a Pro case from a Con case saved the same minute.
+// A side column omits its own side; a both-sides item still says so, since it is listed in both columns.
+function ItemRow({ item, omitSide }: { item: DebateItem; omitSide?: 'pro' | 'con' }) {
   const { t } = useTranslation();
   const openItem = useOpenDebateItem();
   const canOpen = item.kind !== 'crossfire';
+  const showSide = item.side && item.side !== 'neutral' && item.side !== omitSide;
   return (
     <li className="flex items-center justify-between gap-3 border-b border-border/60 py-2 last:border-b-0">
       <div className="min-w-0 flex-1">
         <p className="truncate text-sm font-medium">{item.title}</p>
         <p className="text-xs text-muted-foreground">
-          {t(`debate.kinds.${item.kind}`)} · {formatEuropeanDateTime(item.createdAt)}
+          {t(`debate.kinds.${item.kind}`)}
+          {showSide && <> · {t(sideLabelKey(item.side!))}</>}
+          {' · '}{formatEuropeanDateTime(item.createdAt)}
         </p>
       </div>
       {/* The name starts with the visible word (WCAG 2.5.3) and ends with which item it loads. */}
       {canOpen && (
         <Button variant="ghost" size="sm" className="shrink-0" onClick={() => openItem(item)}>
-          {t('historyPage.load', 'Load')}<span className="sr-only"> {item.title}</span>
+          {t('historyPage.load', 'Load')}<span className="sr-only"> {item.title}, {t(`debate.kinds.${item.kind}`)}{item.side && item.side !== 'neutral' ? `, ${t(sideLabelKey(item.side))}` : ''}</span>
         </Button>
       )}
     </li>
@@ -62,7 +67,7 @@ function SideSection({ side, items }: { side: 'pro' | 'con'; items: DebateItem[]
       <CardContent>
         {relevant.length === 0
           ? <p className="text-sm text-muted-foreground">{t('debate.sideEmpty', 'No work for this side yet.')}</p>
-          : <ul>{relevant.slice(0, 4).map((item) => <ItemRow key={`${item.kind}-${item.id}`} item={item} />)}</ul>}
+          : <ul>{relevant.slice(0, 4).map((item) => <ItemRow key={`${item.kind}-${item.id}`} item={item} omitSide={side} />)}</ul>}
         <ViewAll to="/app/history" count={relevant.length} shown={4} />
       </CardContent>
     </Card>