feat(tui): RLPH single-queue repeating queue/steer + governance docs

[ARCJ137442]

Summary

This PR delivers the current Exomind phase for Idea 1 + Idea 2 with issue-driven tracking.

Idea 1 (Doc)

• Added implementation proposal document:
  • docs/exomind-norm-governance-plan.md
• Added multi-step task tracking document:
  • docs/exomind-task-list.md

Idea 2 (RLPH / TUI)

• Implemented buffered queue channel handling (Ctrl+Tab / Shift+Tab) path integration.
• Implemented single shared queue behavior for normal + repeating messages.
• Added repeating message rotation: repeating items re-enter queue tail after dispatch.
• Added support for both repeating types:
  • repeating queue message
  • repeating steer message
• Added preview metadata and UI differentiation in PendingInputPreview:
  • normal queue
  • repeating queue
  • normal steer
  • repeating steer
• Kept existing Alt+Up queue-edit workflow reusable through shared queue model.

Test/Build updates

• Updated affected tests/constructors for new fields and InputResult::BufferedQueued match exhaustiveness.

Verification

• cargo check -p codex-tui : ✅ passed
• cargo test -p codex-tui --lib --no-run : ⚠ failed in Windows linker stage due environment disk-space/link constraints (LNK1106, os error 112), not Rust type errors.

Tracking

• Refs Idea 1: 规范全量内化与自治治理体系（持续跟踪） #1
• Refs Epic: Idea 2 RLPH 单队列持续重复消息机制（Queue/Steer） #2
• Refs Context Ledger: 需求/计划外部语义化同步账本 #3

[ARCJ137442]

补充验证（2026-03-06）：\n\n- 额外尝试了 CARGO_PROFILE_DEV_DEBUG=0 cargo test -p codex-tui --lib --no-run 以降低链接体积。\n- 结果仍因本机磁盘空间限制失败（aws-lc / linker 阶段 No space left on device, os error 112），不是新的 Rust 类型错误。\n- 为恢复环境已执行：cargo clean -p codex-tui -p codex-cli -p codex-cloud-tasks（释放约 3.0 GiB）。

[ARCJ137442]

验证补充（2026-03-06）：\n\n- cargo test -p codex-tui --lib --no-run 已通过（使用 CARGO_TARGET_DIR=G:\codex-rs-target）。\n- pending_input_preview::tests 已通过。\n- 已提交并推送 snapshot 更新与任务清单更新：cf31fa6da。