fix(analytics): avg duration handles both quoted and unquoted duration_s

The per-skill (avg X) column in gstack-analytics silently rendered as
"0s" for any /skill whose completion events emit "duration_s" as a
quoted JSON string. The completion-status preamble that every skill
runs (scripts/resolvers/preamble/generate-completion-status.ts:70) and
gstack-codex-probe both emit "duration_s":"<N>", so the dominant case
was wrong; gstack-telemetry-log emits the unquoted form which is why
the bug didn't show up everywhere.

The AVG_DUR awk block split each line on the literal "duration_s":
prefix and then ran gsub(/[^0-9.].*/, "", parts[2]). For unquoted
600,"ts":... the gsub matches at the comma and yields "600". For
quoted "600","ts":... it matches at the leading double-quote
(position 0) and the trailing .* consumes the entire rest of the
line, leaving an empty string that never accumulates.

The companion TOTAL_DURATION calculation (line 114) uses a different
parser keyed on -F'[:,]' field splitting + non-greedy gsub, so it
isolates the value before stripping and was always correct. Only the
per-skill average was broken.

Fix: strip an optional leading double-quote with sub(/^"/, ...) before
the existing gsub so quoted and unquoted forms take the same path.

Adds test/gstack-analytics-avg-duration.test.ts covering quoted-only,
unquoted-only, mixed, and total-time aggregation. The test fails on
the pre-fix tree (expected "2m", got "0s") and passes after the fix.

bin/gstack-analytics

@@ -168,7 +168,9 @@ echo "$SKILL_COUNTS" | while read -r COUNT SKILL; do
       # Extract duration_s value using split on "duration_s":
       n = split($0, parts, "\"duration_s\":")
       if (n >= 2) {
-        # parts[2] starts with the value, e.g. "142,"
+        # parts[2] starts with the value: unquoted (142,...) or quoted ("142",...).
+        # Strip a leading double-quote so the digit-extraction below handles both forms.
+        sub(/^"/, "", parts[2])
         gsub(/[^0-9.].*/, "", parts[2])
         if (parts[2]+0 > 0) { total += parts[2]; count++ }
       }

test/gstack-analytics-avg-duration.test.ts

@@ -0,0 +1,87 @@
+import { describe, test, expect, beforeEach, afterEach } from 'bun:test';
+import { spawnSync } from 'child_process';
+import * as fs from 'fs';
+import * as os from 'os';
+import * as path from 'path';
+
+const ROOT = path.resolve(import.meta.dir, '..');
+const BIN = path.join(ROOT, 'bin', 'gstack-analytics');
+
+// Verifies the per-skill avg-duration calculation in the bash dashboard
+// (bin/gstack-analytics) parses both unquoted and quoted "duration_s"
+// values from skill-usage.jsonl. The completion-status preamble (and
+// gstack-codex-probe) emit quoted strings; gstack-telemetry-log emits
+// unquoted numbers. The previous awk regex stripped from the leading
+// quote, silently yielding avg=0 for quoted-only skills.
+describe('gstack-analytics avg duration parsing', () => {
+  let stateDir: string;
+
+  beforeEach(() => {
+    stateDir = fs.mkdtempSync(path.join(os.tmpdir(), 'gstack-analytics-test-'));
+    fs.mkdirSync(path.join(stateDir, 'analytics'), { recursive: true });
+  });
+
+  afterEach(() => {
+    fs.rmSync(stateDir, { recursive: true, force: true });
+  });
+
+  function run(jsonl: string[]): string {
+    fs.writeFileSync(
+      path.join(stateDir, 'analytics', 'skill-usage.jsonl'),
+      jsonl.join('\n') + '\n',
+    );
+    const r = spawnSync('bash', [BIN, 'all'], {
+      env: { ...process.env, GSTACK_STATE_DIR: stateDir },
+      encoding: 'utf-8',
+    });
+    if (r.status !== 0) {
+      throw new Error(`gstack-analytics failed (status ${r.status}): ${r.stderr}`);
+    }
+    return r.stdout;
+  }
+
+  // Pull "(avg <value>)" out of a /skill row.
+  function avgFor(report: string, skill: string): string {
+    const re = new RegExp(`/${skill}\\b[^\\n]*\\(avg ([^)]+)\\)`);
+    const m = report.match(re);
+    if (!m) throw new Error(`no avg row for /${skill} in:\n${report}`);
+    return m[1];
+  }
+
+  test('quoted "duration_s" values are averaged (regression: previously read as 0)', () => {
+    const output = run([
+      '{"skill":"review","duration_s":"120","outcome":"success","ts":"2026-05-25T10:00:00Z"}',
+      '{"skill":"review","duration_s":"180","outcome":"success","ts":"2026-05-25T11:00:00Z"}',
+      '{"skill":"review","duration_s":"60","outcome":"success","ts":"2026-05-25T12:00:00Z"}',
+    ]);
+    // (120+180+60)/3 = 120s = 2m
+    expect(avgFor(output, 'review')).toBe('2m');
+  });
+
+  test('unquoted numeric "duration_s" values are averaged', () => {
+    const output = run([
+      '{"skill":"qa","duration_s":300,"outcome":"success","ts":"2026-05-25T13:00:00Z"}',
+      '{"skill":"qa","duration_s":420,"outcome":"success","ts":"2026-05-25T14:00:00Z"}',
+    ]);
+    // (300+420)/2 = 360s = 6m
+    expect(avgFor(output, 'qa')).toBe('6m');
+  });
+
+  test('mixed quoted and unquoted values average together', () => {
+    const output = run([
+      '{"skill":"ship","duration_s":"30","outcome":"success","ts":"2026-05-25T10:00:00Z"}',
+      '{"skill":"ship","duration_s":50,"outcome":"success","ts":"2026-05-25T11:00:00Z"}',
+    ]);
+    // (30+50)/2 = 40s — under 60 so shown in seconds, exercising both paths.
+    expect(avgFor(output, 'ship')).toBe('40s');
+  });
+
+  test('total time sums quoted and unquoted across all skills', () => {
+    const output = run([
+      '{"skill":"review","duration_s":"120","outcome":"success","ts":"2026-05-25T10:00:00Z"}',
+      '{"skill":"qa","duration_s":300,"outcome":"success","ts":"2026-05-25T13:00:00Z"}',
+    ]);
+    // 120+300 = 420s = 7m
+    expect(output).toContain('Total time: 7m');
+  });
+});