feat(leaderboard): interactive 're-run' button — visitor experiences variance

Hero hook was 'we re-ran the same agent twice, got 0.0 then 0.7'.
Static text. Visitor reads it, moves on.

Replaced with a clickable '↻ Re-run claude-code on tool-001' button
that cycles through 5 real observed scores from the dev-time runs
(0.00 → 0.70 → 0.50 → 0.00 → 0.70). Each click changes the displayed
score with a 180ms fade. Color shifts red ↔ yellow ↔ green by score.

After 2 clicks the hint text changes from 'click. the number changes.
same prompt, same Docker sandbox.' to 'Trial N of 5 observed runs.
This is what variance looks like.'

Why this matters: visitors who PHYSICALLY re-trigger the score swing
internalize the variance message in 3 seconds. Stronger than reading
about it. ~30 LOC of vanilla JS, no deps.

Author: jackjin1997

docs/index.html

@@ -903,13 +903,45 @@
         <span class="pulse"></span> Live Benchmark
       </div>
       <h1 class="reveal reveal-d1">
-        We re-ran the same agent twice.<br/>
-        Got <span style="color: var(--red);">0.0</span> then <span style="color: var(--green);">0.7</span>.
+        Same agent. Same task.<br/>
+        Score: <span id="hero-score" style="color: var(--red); transition: color 0.4s;">0.00</span>
       </h1>
       <p class="subtitle reveal reveal-d2">
         Most AI agent leaderboards report a single number. We don't think that's honest.<br/>
-        This one shows variance.
+        <button id="hero-rerun" style="margin-top: 14px; padding: 10px 20px; background: var(--bg-card); border: 1px solid var(--border-light); border-radius: 8px; color: var(--text); font-size: 0.95em; cursor: pointer; font-family: inherit; font-weight: 600; transition: all 0.2s;">↻ Re-run claude-code on tool-001</button>
+        <span id="hero-rerun-hint" style="display: block; margin-top: 10px; font-size: 0.85em; color: var(--text-secondary);">Click. The number changes. Same prompt, same Docker sandbox.</span>
       </p>
+      <script>
+        // Real per-trial observed scores from Claude Code on tool-001 (see docs/findings.md)
+        // Cycling through them shows visitors what "single-trial" leaderboards hide.
+        (function(){
+          var trials = [0.00, 0.70, 0.50, 0.00, 0.70];
+          var idx = 0;
+          var btn = document.getElementById('hero-rerun');
+          var score = document.getElementById('hero-score');
+          var hint = document.getElementById('hero-rerun-hint');
+          function color(v) {
+            if (v >= 0.6) return 'var(--green)';
+            if (v >= 0.3) return 'var(--yellow)';
+            return 'var(--red)';
+          }
+          btn.addEventListener('click', function(){
+            idx = (idx + 1) % trials.length;
+            var v = trials[idx];
+            score.style.opacity = '0.3';
+            setTimeout(function(){
+              score.textContent = v.toFixed(2);
+              score.style.color = color(v);
+              score.style.opacity = '1';
+            }, 180);
+            if (idx >= 2) {
+              hint.textContent = 'Trial ' + (idx + 1) + ' of 5 observed runs. This is what variance looks like.';
+            }
+          });
+          btn.addEventListener('mouseenter', function(){ btn.style.background = 'var(--accent-glow)'; btn.style.borderColor = 'var(--accent-light)'; });
+          btn.addEventListener('mouseleave', function(){ btn.style.background = 'var(--bg-card)'; btn.style.borderColor = 'var(--border-light)'; });
+        })();
+      </script>
       <div class="hero-actions reveal reveal-d3">
         <a href="https://github.com/jackjin1997/AgentBench-Live" target="_blank" class="btn btn-primary">
           <svg viewBox="0 0 16 16"><path d="M8 0C3.58 0 0 3.58 0 8c0 3.54 2.29 6.53 5.47 7.59.4.07.55-.17.55-.38 0-.19-.01-.82-.01-1.49-2.01.37-2.53-.49-2.69-.94-.09-.23-.48-.94-.82-1.13-.28-.15-.68-.52-.01-.53.63-.01 1.08.58 1.23.82.72 1.21 1.87.87 2.33.66.07-.52.28-.87.51-1.07-1.78-.2-3.64-.89-3.64-3.95 0-.87.31-1.59.82-2.15-.08-.2-.36-1.02.08-2.12 0 0 .67-.21 2.2.82.64-.18 1.32-.27 2-.27.68 0 1.36.09 2 .27 1.53-1.04 2.2-.82 2.2-.82.44 1.1.16 1.92.08 2.12.51.56.82 1.27.82 2.15 0 3.07-1.87 3.75-3.65 3.95.29.25.54.73.54 1.48 0 1.07-.01 1.93-.01 2.2 0 .21.15.46.55.38A8.013 8.013 0 0016 8c0-4.42-3.58-8-8-8z"/></svg>