feat(tui): declutter panels list (name + dim expert count)

[pedrofuentes]

panels rows now show the name prominently with 'N experts' dimmed; description dropped from the cramped row (kept in the wide preview pane + panel detail); ListViewport renders item hints in muted (all list screens), NO_COLOR-safe; Phase 11 polish — feedback #1

[pedrofuentes]

Status: CONDITIONAL

Sentinel Review Report

Ref: feature/tui-panels-declutter → main
Report ID: SENTINEL-1756-2742ec4
Reviewed SHA: 2742ec4
Sentinel ruleset: v1
Reviewed at: 2026-06-26T10:45:00-07:00
Mode: standard
Review depth: Tier 2 (full)
Verdict: CONDITIONAL (reviewed SHA 2742ec4)
Required action: FILE_ISSUES_AND_MERGE
Persisted report: #1756 (comment)

Phase 1 — TDD / Test Evidence

• Tests exist & meaningful: ✅ New tests assert the intended row contract and shared hint styling. Evidence: packages/cli/tests/unit/tui/panels-screen.test.tsx:53-56:

  expect(lastFrame()).not.toContain("Exec Panel"); // description dropped from list row
  expect(lastFrame()).toContain("2 experts"); // expert count in hint
  expect(lastFrame()).toContain("3 experts"); // expert count in hint

  Evidence: packages/cli/tests/unit/tui/list-viewport.test.tsx:293-305:

  const sentinelTheme = { ...theme, muted: (s: string) => `[MUTED:${s}]` };
  ...
  expect(lastFrame()).toContain("[MUTED:3 experts]");

• Test-first history verified: ✅ git log --oneline origin/main..HEAD showed 9abc4e0 test(tui): add failing tests for decluttered panels list before 2742ec4 feat(tui): declutter panels list (name + dim expert count). Commit purity verified: git diff-tree --no-commit-id --name-status -r 9abc4e0 listed only packages/cli/tests/unit/tui/list-viewport.test.tsx and packages/cli/tests/unit/tui/panels-screen.test.tsx; git diff-tree --no-commit-id --name-status -r 2742ec4 listed only packages/cli/src/tui/components/lists/ListViewport.tsx and packages/cli/src/tui/screens/PanelsScreen.tsx.
• RED verified: ✅ In a clean detached worktree at 9abc4e0, npx --yes pnpm@10.30.1 exec vitest run tests/unit/tui/panels-screen.test.tsx tests/unit/tui/list-viewport.test.tsx failed as intended: Test Files 2 failed (2) and Tests 3 failed | 36 passed (39). Representative failure: expected '1/1\n\u001b[7m› Panel Name 3 experts…' to contain '[MUTED:3 experts]'.
• Full suite green on SHA: ✅ In a clean detached worktree at 2742ec425b9306d73f8d086de7ccc104109cb9af, npx --yes pnpm@10.30.1 exec vitest run tests/unit/tui passed Test Files 95 passed (95) / Tests 903 passed (903); npx --yes pnpm@10.30.1 test passed Test Files 358 passed | 1 skipped (359) / Tests 4521 passed | 1 skipped (4522).
• Lint/typecheck: ✅ npx --yes pnpm@10.30.1 lint and npx --yes pnpm@10.30.1 typecheck both exited 0.
• Coverage: ✅ TMPDIR="$PWD/.tmptest" npx --yes pnpm@10.30.1 exec vitest run --coverage exited 0. Summary: statements 91.92%, branches 82.2%, functions 92.44%, lines 93.04% (meets configured thresholds in packages/cli/vitest.config.ts: 90/75/90/90).
• Branch base: ✅ git merge-base origin/main HEAD = 4cf1b8c610c6f3f4e83d0cf692624568c4317add, matching origin/main.
• Scope: ✅ Changed files limited to PanelsScreen.tsx, ListViewport.tsx, and their tests.

Phase 1.5 — Fast-path Evaluation

🔴 count: 0 | LOC: 19 non-test changed lines (≤150: Y) | Security paths: Y (terminal UI renders user/DB-derived panel names) | New deps: N | Commit types qualify: N (feat present)
→ Fast-path eligible: NO → Phase 2

Phase 2 — Execution Log

Dim	Tool Call	Agent ID / Ref	Status
A1	task(agent_type="general-purpose", name="sentinel-pr1756-a1", model="claude-sonnet-4.6", reasoning_effort="medium") in a single multi_tool_use.parallel dispatch	N/A (sync task tool returned no ID)	✅ No findings
A2	task(agent_type="general-purpose", name="sentinel-pr1756-a2", model="claude-sonnet-4.6", reasoning_effort="medium") in the same parallel dispatch	N/A (sync task tool returned no ID)	✅ No findings
B	task(agent_type="general-purpose", name="sentinel-pr1756-b", model="claude-sonnet-4.6", reasoning_effort="medium") in the same parallel dispatch	N/A (sync task tool returned no ID)	✅ No findings
C	task(agent_type="general-purpose", name="sentinel-pr1756-c", model="claude-sonnet-4.6", reasoning_effort="medium") in the same parallel dispatch	N/A (sync task tool returned no ID)	✅ No findings
D	task(agent_type="general-purpose", name="sentinel-pr1756-d", model="claude-sonnet-4.6", reasoning_effort="medium") in the same parallel dispatch	N/A (sync task tool returned no ID)	✅ Two 🟡 findings
E	N/A	N/A	✅ Auto-skipped: no dependency/supply-chain surface changed
F	task(agent_type="general-purpose", name="sentinel-pr1756-f", model="claude-haiku-4.5") in the same parallel dispatch	N/A (sync task tool returned no ID)	✅ No actionable findings
Persist	gh api repos/pedrofuentes/council/issues/1756/comments --method POST/PATCH	#1756 (comment)	✅ Full report posted to PR

Findings

• 🔴 CRITICAL: 0
• 🟡 IMPORTANT: 2 new / 0 known
• 🟢 MINOR: 0

Details (ordered by severity)

1. [🟡 IMPORTANT] Preview-pane test does not prove the description comes from the preview pane — packages/cli/tests/unit/tui/panels-screen.test.tsx:257

   • Evidence:

     expect(lastFrame()).toContain("Finance strategy");

   • Impact: Trigger: a future wide-layout regression reintroduces the panel description into the list row while leaving or removing the preview rendering. Mechanism: this assertion only checks the whole frame for the description, so it cannot distinguish preview text from list-row text. Consequence: the regression could pass the preview preservation test even though the UX contract is that descriptions are omitted from list rows and kept in the wide preview/detail surfaces.
   • Remediation: Add a stronger assertion that binds Finance strategy to preview-only structure, or parse the frame to distinguish the list pane from the bordered preview pane before asserting the description appears there.

2. [🟡 IMPORTANT] Multi-panel count assertions do not bind counts to their panel rows — packages/cli/tests/unit/tui/panels-screen.test.tsx:54-56

   • Evidence:

     expect(lastFrame()).toContain("2 experts"); // expert count in hint
     expect(lastFrame()).toContain("startup-board");
     expect(lastFrame()).toContain("3 experts"); // expert count in hint

   • Impact: Trigger: a future list-item mapping regression swaps or otherwise misassociates expert-count hints between panels. Mechanism: the test checks each panel name/count as independent substrings in the whole frame instead of verifying co-occurrence on the same row. Consequence: a user could see the wrong expert count next to a panel while this multi-panel regression test still passes.
   • Remediation: Assert row-level co-occurrence, e.g. split lastFrame() by newline and require one line containing both acme and 2 experts, and another containing both startup-board and 3 experts.

Follow-ups & Actions

• CONDITIONAL → FILE_ISSUES_AND_MERGE: file issues for both new 🟡 findings (sentinel:important), link them in PR feat(tui): declutter panels list (name + dim expert count) #1756, then merge.
• ⚠️ Do NOT fix 🟡 findings in this PR — file as issues only.
• Only 🟡/🟢 findings remain; there are no 🔴 blockers.

Decision rationale

• TDD choreography is compliant: failing test commit preceded implementation commit, and per-commit diffs are pure.
• The required sanitization invariant is satisfied: PanelsScreen uses toSingleLineDisplay(panel.name) for the panel-name label, and ListViewport also sanitizes item.label and optional item.hint before rendering; the new hint is computed from numeric memberCount.
• Shared ListViewport behavior remains covered: TUI suite, package suite, lint, typecheck, and coverage all pass on the reviewed SHA.
• Phase 2 found no security, resilience, performance, dependency, or documentation blockers.
• Two new 🟡 test-quality gaps remain and must be tracked before merge; therefore the verdict is CONDITIONAL, SHA-bound to 2742ec425b9306d73f8d086de7ccc104109cb9af.

[pedrofuentes]

Sentinel SENTINEL-1756-2742ec4 → CONDITIONAL (0 🔴). Follow-ups filed: #1762, #1763 (🟡 test-quality). Merging per CONDITIONAL policy.