docs(governance): Round 68 panel — full-doc drift sync

johnteee · johnteee · commit 9f88305cc89d · 2026-06-25T13:50:11.000+08:00
Six-lens unanimous pass: sync stale README panel round refs (1–20 → 1–68),
fix METHODOLOGY_MAP.zh-TW nine-skill count, add root North Star/governance
pointers, refresh QUALITY_GATES Phase 2 maintenance narrative, align
make-all claims in AGENTS/GLOSSARY playbook, add test_readme_governance.py.
diff --git a/README.md b/README.md
@@ -10,6 +10,20 @@ Core principle:
 Doing the right thing > doing things right.
 ```
 
+
+
+## North Star
+
+TeaPrompt helps humans and host agents choose the **right amount of rigor** for a task, record **why** decisions were made, and verify outcomes with **evidence** — using composable prompt layers and nine workflow skills as natural-language harness policy, **without** operating its own agent runtime.
+
+Full library docs: [reflective-prompt-library/README.md](reflective-prompt-library/README.md).
+
+## Governance
+
+- **Contributing:** [CONTRIBUTING.md](CONTRIBUTING.md) — quality gates, routing maintenance (R8–R12), `make all`
+- **Panel record:** [multi-agent-panel-consensus](reflective-prompt-library/plans/multi-agent-panel-consensus-2026-06-25.md) — six-lens Socratic consensus (Rounds 1–68)
+- **Operator playbook:** [GLOSSARY.md](reflective-prompt-library/GLOSSARY.md) — Governance Maintenance Playbook
+
 The repository contains:
 
 - `reflective-prompt-library/`: reusable Markdown prompts for thinking, planning, implementation, review, research, risk handling, and context handoff.
diff --git a/README.zh-TW.md b/README.zh-TW.md
@@ -12,6 +12,20 @@ TeaPrompt 是一套給 AI 協作工程使用的反思型 Prompt 與 Workflow Lib
 Doing the right thing > doing things right.
 ```
 
+
+
+## 北極星（North Star）
+
+TeaPrompt 幫助人類與宿主 agent 為任務選擇**恰當的嚴謹度**，記錄**決策理由**，並以**證據**驗證結果 —— 透過可組合的 prompt 層與九個 workflow skills 作為自然語言 harness policy，**不**自建 agent runtime。
+
+完整庫文件：[reflective-prompt-library/README.zh-TW.md](reflective-prompt-library/README.zh-TW.md)。
+
+## 治理與貢獻
+
+- **貢獻指南：** [CONTRIBUTING.md](CONTRIBUTING.md)（英文）— 品質閘門、路由維護、`make all`
+- **多視角共識紀錄：** [multi-agent-panel-consensus](reflective-prompt-library/plans/multi-agent-panel-consensus-2026-06-25.md)（Rounds 1–68）
+- **維護手冊：** [GLOSSARY.md](reflective-prompt-library/GLOSSARY.md) — Governance Maintenance Playbook
+
 此 repository 包含：
 
 - `reflective-prompt-library/`：可重用的提示詞文件（思考、規劃、實作、審查、研究、風險、交接）。
diff --git a/reflective-prompt-library/06-repo/AGENTS.md b/reflective-prompt-library/06-repo/AGENTS.md
@@ -29,7 +29,7 @@ runtime. For non-trivial routing, governance, or skill-selection work:
    `reflective-research` (single host, persona lenses); judgment record at
    [plans/multi-agent-panel-consensus-2026-06-25.md](../plans/multi-agent-panel-consensus-2026-06-25.md).
 
-Run `make validate` before claiming routing or governance changes are verified.
+Run `make all` before claiming routing or governance changes are verified.
 
 ---
 
diff --git a/reflective-prompt-library/GLOSSARY.md b/reflective-prompt-library/GLOSSARY.md
@@ -337,9 +337,9 @@ Curated top-of-cheatsheet summary of high-confusion routing traps (ROUTE-002 hol
 
 ## Governance Maintenance Playbook / 治理維護手冊
 
-Ongoing upkeep after panel close (Rounds 1–67). Not agent instructions — operator checklist.
+Ongoing upkeep after panel close (Rounds 1–68). Not agent instructions — operator checklist.
 
-**Operational test:** Before router tuning, add fresh ROUTE-002/003 holdout phrases; run `make validate`; record decisions in `PROJECT_KNOWLEDGE.md` Decision Index when governance surface changes.
+**Operational test:** Before router tuning, add fresh ROUTE-002/003 holdout phrases; run `make all`; record decisions in `PROJECT_KNOWLEDGE.md` Decision Index when governance surface changes.
 
 1. Run `make all` on every governance/routing change.
 2. Add holdout cases **before** tuning `route_paraphrase_eval.py` keyword rules; `validate_route_fixture.py` blocks accidental shrinkage.
diff --git a/reflective-prompt-library/METHODOLOGY_MAP.zh-TW.md b/reflective-prompt-library/METHODOLOGY_MAP.zh-TW.md
@@ -121,6 +121,6 @@ NLAH 論文（明確引用 `SKILL.md` 類型檔案作為直接靈感來源）提
 | 工作流與智能體層 | 契合 | 完成 | 工作流引擎、計畫與記憶整合 (`04-agent/`)。 |
 | 領域套件層 | 契合 | 完成 | 保留策略提示詞於 `05-domain/`。 |
 | 專案模板層 | 契合 | 完成 | 維持 `AGENTS.md` 和 `cursor-rules.md` 模板於 `06-repo/`。 |
-| 技能與動作層 | 契合 | 擴充 | 維持 8 個生命週期技能，並加入窄範圍 minimality gate 以處理反膨脹決策。 |
+| 技能與動作層 | 契合 | 擴充 | 維持 9 個凍結 workflow skills（含 reflective-minimality gate）以處理反膨脹決策。 |
 | 品質閘門與驗證 | 契合 | 完成 | 標準化技能層級的品質檢查。 |
 | 治理與能力風險 | 契合 | 完成 | 使用 `reflective-risk` 處理高風險邊界。 |
diff --git a/reflective-prompt-library/PROJECT_KNOWLEDGE.md b/reflective-prompt-library/PROJECT_KNOWLEDGE.md
@@ -41,7 +41,7 @@ operating its own agent runtime. See [README.md](README.md#north-star).
 > does not silently expand. Completed milestones are retired from this section;
 > their outcomes live in the Decision Index.
 
-**Ongoing (post Round 67 panel):** maintain ROUTE-001/002/003 evals, governance validators,
+**Ongoing (post Round 68 panel):** maintain ROUTE-001/002/003 evals, governance validators,
 and holdout expansion before router tuning — see [GLOSSARY.md](GLOSSARY.md) Governance
 Maintenance Playbook. Post-panel maintenance (e.g. ROUTING_CONTRACT R11 approved-spec delivery)
 follows the same holdout-before-tune rule. No open implementation blockers from Rounds 8–20;
diff --git a/reflective-prompt-library/README.md b/reflective-prompt-library/README.md
@@ -30,7 +30,7 @@ Pick **Strictness L1–L6** first (`skills/reflective-dispatch/SKILL.md`, [GLOSS
 
 ## Governance Panel Record
 
-Multi-agent Socratic consensus on project goals and the nine skills (Rounds 1–20, options A–AJ) is recorded in [plans/multi-agent-panel-consensus-2026-06-25.md](plans/multi-agent-panel-consensus-2026-06-25.md). Run `make validate` before claiming routing or governance changes.
+Multi-agent Socratic consensus on project goals and the nine skills (Rounds 1–68, options A–CZ) is recorded in [plans/multi-agent-panel-consensus-2026-06-25.md](plans/multi-agent-panel-consensus-2026-06-25.md). Run `make all` before claiming routing or governance changes are verified.
 
 ## Directory Map
 
diff --git a/reflective-prompt-library/plans/QUALITY_GATES_SUMMARY.md b/reflective-prompt-library/plans/QUALITY_GATES_SUMMARY.md
@@ -306,14 +306,15 @@ The latest router improvement uses concept-level boundary rules for these cases
 
 ROUTE-002 measures unseen phrasing separately from ROUTE-001. Round 7 (2026-06-25) added Traditional Chinese holdout groups with matching router intent keywords — fairness test without full `SKILL.md` translation. Round 65 expanded to 36 groups / 100 phrases; post-panel maintenance added ROUTING_CONTRACT **R11** (approved-spec delivery) with ROUTE-003 `implement_not_plan_trap` at 100%. Treat this as a seeded holdout, not proof of broad semantic routing; add fresh cases before further router tuning.
 
-## Phase 2 Status (Round 21 audit)
+## Phase 2 Status (post-Round 68 maintenance)
 
 ### Done ✅
 
 1. **CI/CD** — `.github/workflows/python-tools.yml` runs `make all` on push/PR
 2. **ROUTE-001/002/003 in CI** — 128 + 100 + 31 paraphrases at 100% consistency (seeded fixtures); `validate_route_fixture.py` gates minimum coverage
 3. **Governance validators** — links, lint, governance metadata, PROJECT_KNOWLEDGE, benchmark fixture, skill examples
 4. **Harness policy docs** — CONTRIBUTING, AGENTS, SKILL_INSTALLATION, maintenance playbook
+5. **Doc anti-drift** — `test_routing_contract.py`, cheatsheet parity tests, `test_readme_governance.py` (170+ pytest in CI)
 
 ### Ongoing maintenance (not blockers)
 
@@ -374,7 +375,7 @@ python3 reflective-prompt-library/plans/route_paraphrase_eval.py
 
 ## Conclusion
 
-Phase 1 quality-gate tooling and documentation are **complete**. Routing consistency on seeded fixtures (ROUTE-001 tuning, ROUTE-002 holdout, ROUTE-003 adversarial) is at **100%** as of Round 65 plus R11 approved-spec delivery maintenance; treat this as regression protection, not proof of broad semantic routing. TeaPrompt has:
+Phase 1 quality-gate tooling and documentation are **complete**. Routing consistency on seeded fixtures (ROUTE-001 tuning, ROUTE-002 holdout, ROUTE-003 adversarial) is at **100%** as of Round 68 (R11 approved-spec delivery + R12 boundary quick-cue maintenance); treat this as regression protection, not proof of broad semantic routing. TeaPrompt has:
 
 - ✅ Automated validation to prevent quality degradation
 - ✅ Machine-readable index for tool integration
@@ -383,4 +384,4 @@ Phase 1 quality-gate tooling and documentation are **complete**. Routing consist
 - ✅ Benchmark fixture gate plus optional manual benchmark runs
 - ✅ Research-backed design decisions
 
-The project is positioned to grow sustainably with quality discipline built in from the start. **No open implementation blockers** remain from the Rounds 8–20 panel work (Round 21 audit). The next measurable quality target is **holdout expansion before router tuning** and optional manual baseline-vs-skill benchmark runs — not shipping new core skills without promotion evidence.
+The project is positioned to grow sustainably with quality discipline built in from the start. **No open implementation blockers** remain from panel Rounds 1–68; work is recurrence-gated maintenance per playbook. The next measurable quality target is **holdout expansion before router tuning** and optional manual baseline-vs-skill benchmark runs — not shipping new core skills without promotion evidence.
diff --git a/reflective-prompt-library/plans/multi-agent-panel-consensus-2026-06-25.md b/reflective-prompt-library/plans/multi-agent-panel-consensus-2026-06-25.md
@@ -1355,9 +1355,123 @@ User directive: **do anything you want** — reopen panel for post-Round-66 oper
 
 - `make all`: pytest + ROUTE-001/002/003 100%
 
+## Round 68 — Full-doc drift sync (2026-06-25)
+
+User directive: **review all docs** — reopen six-lens panel for cross-surface drift after Round 67.
+
+### CS: Sync library README panel record?
+
+| Lens | Position |
+| --- | --- |
+| Opus | **Agree** — "Rounds 1–20" stale; update to Rounds 1–68, options A–CZ |
+| Codex | **Agree** — `make all` for verified claims, not `make validate` alone |
+| Gemini | **Agree** — entry README is governance discoverability surface |
+| Composer | **Agree** — anti-drift via `test_readme_governance.py` |
+| Sakana | **Agree** — panel record is historical truth, must track seal round |
+| GLM | **Agree** — minimal one-line fix |
+
+**Consensus:** **Agree** — `reflective-prompt-library/README.md` Governance Panel Record.
+
+### CT: Fix METHODOLOGY_MAP.zh-TW skill count?
+
+| Lens | Position |
+| --- | --- |
+| All six | **Agree** — "8 個生命週期技能" contradicts frozen nine-skill policy |
+
+**Consensus:** **Agree** — update to nine frozen workflow skills.
+
+### CU: Root README North Star + governance pointers?
+
+| Lens | Position |
+| --- | --- |
+| Opus | **Agree** — root is fork entry; link CONTRIBUTING + panel + playbook |
+| Codex | **Agree** — brief North Star, link to library README for depth |
+| Gemini | **Agree** — README.zh-TW parity (北極星 + 治理) |
+| Composer | **Agree** — do not duplicate full skill tables at root |
+| Sakana | **Reject** duplicating panel methodology essay at root |
+| GLM | **Agree** — pointers only |
+
+**Consensus:** **Agree** — root `README.md` + `README.zh-TW.md` North Star and governance sections.
+
+### CV: Refresh QUALITY_GATES Phase 2 section?
+
+| Lens | Position |
+| --- | --- |
+| Opus | **Agree** — "Round 21 audit" header stale after R12 + Round 67 |
+| Codex | **Agree** — add doc anti-drift pytest bullet; restore if corrupted |
+| Gemini | **Agree** — conclusion should say maintenance mode, not blockers |
+| Composer | **Agree** — `test_quality_gates_summary.py` guards Phase 2 header |
+| Sakana | **Reject** re-litigating Phase 1 completeness |
+| GLM | **Agree** — metrics unchanged; narrative only |
+
+**Consensus:** **Agree** — Phase 2 → post-Round 68 maintenance; anti-drift test mention.
+
+### CW: `test_readme_governance.py`?
+
+| Lens | Position |
+| --- | --- |
+| All six | **Agree** — guard README round refs, root governance, zh-TW skill count |
+
+**Consensus:** **Agree** — new pytest module.
+
+### CX: Expand ROUTE holdout / tune router?
+
+| Lens | Position |
+| --- | --- |
+| All six | **Reject** — ROUTE-001/002/003 remain 100% |
+
+**Consensus:** **Reject** — no router changes.
+
+### CY: GLOSSARY playbook `make all` + round bump?
+
+| Lens | Position |
+| --- | --- |
+| All six | **Agree** — operational test line must match playbook step 1 |
+
+**Consensus:** **Agree** — Rounds 1–68; `make all` in operational test.
+
+### CZ: Final seal?
+
+| Lens | Position |
+| --- | --- |
+| All six | **Agree** — reseal at Round 68 |
+
+**Consensus:** **Agree close** — panel resealed Round 68.
+
+### Round 68 verdict table
+
+| ID | Option | Verdict | Action |
+| --- | --- | --- | --- |
+| CS | Library README panel sync | **Agree** | Rounds 1–68, `make all` |
+| CT | zh-TW methodology map | **Agree** | Nine skills |
+| CU | Root README governance | **Agree** | EN + zh-TW North Star |
+| CV | QUALITY_GATES Phase 2 | **Agree** | Maintenance narrative |
+| CW | README governance pytest | **Agree** | `test_readme_governance.py` |
+| CX | Holdout/router expansion | **Reject** | — |
+| CY | Playbook make all | **Agree** | `GLOSSARY.md` |
+| CZ | Final seal | **Agree** | Panel resealed Round 68 |
+
+**All roles agree.**
+
+## Implemented Changes (Round 68)
+
+- `reflective-prompt-library/README.md`: Governance Panel Record Rounds 1–68
+- `METHODOLOGY_MAP.zh-TW.md`: nine frozen workflow skills
+- `README.md`, `README.zh-TW.md`: North Star + governance pointers
+- `QUALITY_GATES_SUMMARY.md`: Phase 2 post-Round 68 maintenance narrative
+- `GLOSSARY.md`: playbook Rounds 1–68; operational test `make all`
+- `06-repo/AGENTS.md`: `make all` for verified claims
+- `plans/tests/test_readme_governance.py`: README + zh-TW anti-drift
+- `test_glossary_structure.py`, `test_quality_gates_summary.py`: Round 68 guards
+- `PROJECT_KNOWLEDGE.md`: Decision Index entry
+
+## Verification (Round 68)
+
+- `make all`: pytest + ROUTE-001/002/003 100%
+
 ## Panel status (updated)
 
-**Sealed 2026-06-25** after **Round 67** (options A–CR). Contributor/operator doc parity complete. Future work remains recurrence-gated maintenance per playbook.
+**Sealed 2026-06-25** after **Round 68** (options A–CZ). Full-doc drift sync complete. Future work remains recurrence-gated maintenance per playbook.
 
 _Ongoing maintenance: [GLOSSARY.md](../GLOSSARY.md) Governance Maintenance Playbook — expand ROUTE-002/003 holdout before `route_paraphrase_eval.py` tuning (R8–R12)._
 
diff --git a/reflective-prompt-library/plans/tests/test_glossary_structure.py b/reflective-prompt-library/plans/tests/test_glossary_structure.py
@@ -30,10 +30,10 @@ def test_round_boundary_terms_present(glossary_text: str):
         assert heading in glossary_text, f"missing glossary section: {heading}"
 
 
-def test_maintenance_playbook_references_round_67(glossary_text: str):
+def test_maintenance_playbook_references_round_68(glossary_text: str):
     playbook = glossary_text.split("## Governance Maintenance Playbook", 1)[1]
-    assert "Rounds 1–67" in playbook or "Rounds 1-67" in playbook
-    assert "Rounds 1–66" not in playbook and "Rounds 1-66" not in playbook
+    assert "Rounds 1–68" in playbook or "Rounds 1-68" in playbook
+    assert "Rounds 1–67" not in playbook and "Rounds 1-67" not in playbook
 
 
 def test_approved_spec_delivery_separated_from_playbook(glossary_text: str):
diff --git a/reflective-prompt-library/plans/tests/test_quality_gates_summary.py b/reflective-prompt-library/plans/tests/test_quality_gates_summary.py
@@ -94,3 +94,8 @@ def test_route_fixture_gate_mentions_cheatsheet_parity_tests(summary_text: str):
     assert "test_cheatsheet_boundary_quick_cues.py" in section
     assert "test_cheatsheet_" in section and "_parity.py" in section
 
+def test_phase2_status_references_post_round_68_maintenance(summary_text: str):
+    assert "## Phase 2 Status (post-Round 68 maintenance)" in summary_text
+    assert "Round 21 audit" not in summary_text
+    assert "test_readme_governance.py" in summary_text
+
diff --git a/reflective-prompt-library/plans/tests/test_readme_governance.py b/reflective-prompt-library/plans/tests/test_readme_governance.py
@@ -0,0 +1,54 @@
+"""Anti-drift tests for README governance surface and zh-TW skill count."""
+
+from pathlib import Path
+
+import pytest
+
+LIBRARY_README = Path(__file__).parent.parent.parent / "README.md"
+ROOT_README = Path(__file__).parent.parent.parent.parent / "README.md"
+METHODOLOGY_MAP_ZH = Path(__file__).parent.parent.parent / "METHODOLOGY_MAP.zh-TW.md"
+
+CURRENT_PANEL_ROUND = "68"
+CURRENT_PANEL_OPTIONS = "A–CZ"
+
+
+@pytest.fixture(scope="module")
+def library_readme_text() -> str:
+    assert LIBRARY_README.is_file(), f"missing {LIBRARY_README}"
+    return LIBRARY_README.read_text(encoding="utf-8")
+
+
+@pytest.fixture(scope="module")
+def root_readme_text() -> str:
+    assert ROOT_README.is_file(), f"missing {ROOT_README}"
+    return ROOT_README.read_text(encoding="utf-8")
+
+
+@pytest.fixture(scope="module")
+def methodology_map_zh_text() -> str:
+    assert METHODOLOGY_MAP_ZH.is_file(), f"missing {METHODOLOGY_MAP_ZH}"
+    return METHODOLOGY_MAP_ZH.read_text(encoding="utf-8")
+
+
+def test_library_readme_panel_record_current(library_readme_text: str):
+    assert f"Rounds 1–{CURRENT_PANEL_ROUND}" in library_readme_text
+    assert f"options {CURRENT_PANEL_OPTIONS}" in library_readme_text
+    assert "Rounds 1–20" not in library_readme_text
+    assert "options A–AJ" not in library_readme_text
+
+
+def test_library_readme_requires_make_all_for_claims(library_readme_text: str):
+    section = library_readme_text.split("## Governance Panel Record", 1)[1].split("##", 1)[0]
+    assert "`make all`" in section
+
+
+def test_root_readme_has_north_star_and_governance(root_readme_text: str):
+    assert "## North Star" in root_readme_text
+    assert "## Governance" in root_readme_text
+    assert "CONTRIBUTING.md" in root_readme_text
+    assert f"Rounds 1–{CURRENT_PANEL_ROUND}" in root_readme_text
+
+
+def test_methodology_map_zh_tw_lists_nine_skills(methodology_map_zh_text: str):
+    assert "9 個" in methodology_map_zh_text
+    assert "8 個生命週期技能" not in methodology_map_zh_text
