Round 86: composable Human Review preamble guards

johnteee · johnteee · commit a72bf182d61e · 2026-06-25T16:38:57.000+08:00
Add parametrized Human Review preamble guards for composable prompts
(02-engineering–06-repo) via prompt_eval_helpers exact-heading detection;
align HR copy to route through reflective-risk; backfill GLOSSARY playbook
step 17 (R85) and add step 18; governance sync for panel round 86.
diff --git a/README.md b/README.md
@@ -21,7 +21,7 @@ Full library docs: [reflective-prompt-library/README.md](reflective-prompt-libra
 ## 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 1–85)
+- **Panel record:** [multi-agent-panel-consensus](reflective-prompt-library/plans/multi-agent-panel-consensus-2026-06-25.md) — six-lens Socratic consensus (Rounds 1–86)
 - **Operator playbook:** [GLOSSARY.md](reflective-prompt-library/GLOSSARY.md) — Governance Maintenance Playbook
 
 The repository contains:
diff --git a/reflective-prompt-library/03-context/small-context.md b/reflective-prompt-library/03-context/small-context.md
@@ -22,7 +22,7 @@ State what evidence would require escalating to `medium-context.md`.
 
 ## Human Review
 
-Escalate when window limits would hide safety-critical unknowns.
+Escalate to `reflective-risk` when window limits would hide safety-critical unknowns.
 
 ```markdown
 你在小 context window 中工作。請極度節省 token。
diff --git a/reflective-prompt-library/04-agent/memory-consolidation.md b/reflective-prompt-library/04-agent/memory-consolidation.md
@@ -23,7 +23,7 @@ State what would disprove a pattern's claimed recurrence.
 
 ## Human Review
 
-Require human approval before promoting executable knowledge (skills, hooks, runners).
+Escalate to `reflective-risk` before promoting executable knowledge (skills, hooks, runners).
 
 
 ```markdown
diff --git a/reflective-prompt-library/04-agent/retro.md b/reflective-prompt-library/04-agent/retro.md
@@ -23,7 +23,7 @@ Name one proposed rule that would be rejected if the failure cannot recur.
 
 ## Human Review
 
-Escalate promotion candidates that would change agent operating rules to explicit human approval.
+Escalate to `reflective-risk` when promotion candidates would change agent operating rules.
 
 
 ```markdown
diff --git a/reflective-prompt-library/05-domain/creative-template.md b/reflective-prompt-library/05-domain/creative-template.md
@@ -22,7 +22,7 @@ State what preview or validation result would reject the creative spec.
 
 ## Human Review
 
-Require human approval before publishing creative output that affects brand or external audiences.
+Escalate to `reflective-risk` before publishing creative output that affects brand or external audiences.
 
 ```markdown
 你是 Creative Template Architect。請把自然語言創意轉成可驗證、可重播、可約束的 creative spec。
diff --git a/reflective-prompt-library/06-repo/AGENTS.md b/reflective-prompt-library/06-repo/AGENTS.md
@@ -18,6 +18,10 @@ Repository-level harness policy for reflective engineering agents. Primary workf
 
 Name one task that would fail if this harness policy were ignored at session start.
 
+## Human Review
+
+Escalate to `reflective-risk` for irreversible or high-blast-radius work; see **Human Review Required** below for category triggers.
+
 ## Mission
 
 This repository uses Reflective Engineering Agent Protocol.
diff --git a/reflective-prompt-library/06-repo/PROJECT_KNOWLEDGE.template.md b/reflective-prompt-library/06-repo/PROJECT_KNOWLEDGE.template.md
@@ -20,7 +20,7 @@ State what evidence would retire a recorded principle as stale.
 
 ## Human Review
 
-Require human approval before promoting lessons into executable skills or overriding `AGENTS.md`.
+Escalate to `reflective-risk` before promoting lessons into executable skills or overriding `AGENTS.md`.
 
 
 # Project Knowledge — [Project Name]
diff --git a/reflective-prompt-library/06-repo/cursor-rules.md b/reflective-prompt-library/06-repo/cursor-rules.md
@@ -20,7 +20,7 @@ State what test or diff review would reject the change as out of scope.
 
 ## Human Review
 
-Require human approval before editing credential handling, schema, billing, public API, or destructive operations.
+Escalate to `reflective-risk` before editing credential handling, schema, billing, public API, or destructive operations.
 
 ```markdown
 You are working inside an IDE. Prioritize small, safe, reviewable edits.
diff --git a/reflective-prompt-library/GLOSSARY.md b/reflective-prompt-library/GLOSSARY.md
@@ -337,7 +337,7 @@ 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 1–85). Not agent instructions — operator checklist.
+Ongoing upkeep after panel close (Rounds 1–86). Not agent instructions — operator checklist.
 
 **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.
 
@@ -357,3 +357,4 @@ Ongoing upkeep after panel close (Rounds 1–85). Not agent instructions — ope
 14. When editing `01-thinking/` Purpose preambles, keep `Primary workflow surfaces` aligned exactly with `SKILL_THINKING_SOURCES` via `test_thinking_lens_primary_surfaces_match_consumer_graph`; put escalations and adjacent workflow notes in Scope or Human Review, not on the primary line.
 15. When editing composable prompts (`02-engineering`–`06-repo`), keep `Primary workflow surface(s)` aligned with `*_SKILL_LINKS` in `test_prompt_cross_links.py`; use Supporting lens for cross-cutting lenses like `runtime-trust-boundary.md`; put escalate/pair notes in Scope.
 16. When editing `00-core/` prompts, keep `Primary workflow surface(s)` aligned with `CORE_SKILL_LINKS` in `test_prompt_cross_links.py`; put pair/escalation skills in Scope or Human Review, not on the primary line.
+17. When editing composable prompts (`00-core`–`06-repo`), keep `Primary workflow surface(s)` / Supporting-lens preamble lines and run `test_*_prompts_eval_harness.py` primary-surface guards.18. When adding or editing composable prompts (`02-engineering`–`06-repo`) with `## Human Review`, keep preamble escalation routed to `reflective-risk` and run Human Review guards in `test_*_prompts_eval_harness.py` (exact heading match via `prompt_eval_helpers.py`).
diff --git a/reflective-prompt-library/PROJECT_KNOWLEDGE.md b/reflective-prompt-library/PROJECT_KNOWLEDGE.md
@@ -73,6 +73,7 @@ deferred promotions are recurrence-gated — see [panel backlog](plans/multi-age
 ## Decision Index
 
 - 2026-06-25 Round 85 panel — composable prompt Primary workflow surface preamble guards (`test_*_prompts_eval_harness.py`) + Supporting-lens exemption → [record](plans/multi-agent-panel-consensus-2026-06-25.md)
+- 2026-06-25 Round 86 panel — composable Human Review preamble guards + `reflective-risk` routing alignment → [record](plans/multi-agent-panel-consensus-2026-06-25.md)
 - 2026-06-25 Round 84 panel — `00-core` Primary workflow surface parity + primary-line trim → [record](plans/multi-agent-panel-consensus-2026-06-25.md)
 - 2026-06-25 Round 83 panel — composable prompt Primary workflow surface parity (`02-engineering`–`06-repo`) + supporting-lens exemption → [record](plans/multi-agent-panel-consensus-2026-06-25.md)
 > Pointers to the causal trail — plans, reflections, tests, commits. Detail is
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 1–85, options A–FF) 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.
+Multi-agent Socratic consensus on project goals and the nine skills (Rounds 1–86, options A–FJ) 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
@@ -314,7 +314,7 @@ ROUTE-002 measures unseen phrasing separately from ROUTE-001. Round 7 (2026-06-2
 2. **ROUTE-001/002/003 in CI** — 128 + 102 + 53 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`, `test_thinking_prompts_eval_harness.py`, `test_engineering_prompts_eval_harness.py`, `test_prompt_cross_links.py`, `test_core_prompts_eval_harness.py`, `test_agent_prompts_eval_harness.py`, `test_context_prompts_eval_harness.py`, `test_domain_prompts_eval_harness.py`, `test_repo_prompts_eval_harness.py`, `test_validate_governance.py`, `test_validate_links.py`, `test_lint_skills.py`, `test_skill_module_contract.py` (Escalation subsection + Trigger/Methods/Output/Never; 530+ pytest anti-drift suite in CI); reciprocal thinking-lens ↔ skill checks and `00-core` + composable `Primary workflow surface(s)` ↔ `*_SKILL_LINKS` parity in `test_prompt_cross_links.py` (including strict Primary workflow surfaces parity via `test_thinking_lens_primary_surfaces_match_consumer_graph`); Human Review + Escalation route-target guards in thinking/skill contract tests; composable `Primary workflow surface(s)` / Supporting-lens preamble guards in `test_*_prompts_eval_harness.py`
+5. **Doc anti-drift** — `test_routing_contract.py`, cheatsheet parity tests, `test_readme_governance.py`, `test_thinking_prompts_eval_harness.py`, `test_engineering_prompts_eval_harness.py`, `test_prompt_cross_links.py`, `test_core_prompts_eval_harness.py`, `test_agent_prompts_eval_harness.py`, `test_context_prompts_eval_harness.py`, `test_domain_prompts_eval_harness.py`, `test_repo_prompts_eval_harness.py`, `test_validate_governance.py`, `test_validate_links.py`, `test_lint_skills.py`, `test_skill_module_contract.py` (Escalation subsection + Trigger/Methods/Output/Never; 550+ pytest anti-drift suite in CI); reciprocal thinking-lens ↔ skill checks and `00-core` + composable `Primary workflow surface(s)` ↔ `*_SKILL_LINKS` parity in `test_prompt_cross_links.py` (including strict Primary workflow surfaces parity via `test_thinking_lens_primary_surfaces_match_consumer_graph`); Human Review + Escalation route-target guards in thinking/skill contract tests; composable `Primary workflow surface(s)` / Supporting-lens preamble guards and composable `## Human Review` preamble guards (route to `reflective-risk`) in `test_*_prompts_eval_harness.py`
 
 ### Ongoing maintenance (not blockers)
 
@@ -384,4 +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 panel Rounds 1–85; 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.
+The project is positioned to grow sustainably with quality discipline built in from the start. **No open implementation blockers** remain from panel Rounds 1–86; 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
@@ -2364,3 +2364,49 @@ User directive (repeat): review prompts, plans, skills, and Socratic/critical-th
 
 **Resealed 2026-06-25** after **Round 85** (options FC–FF). Composable prompts now have eval_harness preamble guards matching thinking-lens pattern; full library parity (graph + preamble) closed. Holdout expansion remains recurrence-gated maintenance.
 
+
+---
+
+## Round 86 — Composable Human Review preamble guards (2026-06-25)
+
+**Options FG–FJ** | Six-lens panel (Opus, Codex, Gemini, Composer, Sakana, GLM)
+
+### Round 86 options
+
+| ID | Proposal | Verdict |
+| --- | --- | --- |
+| FG | `## Human Review` preamble guards on composable prompts (`02-engineering`–`06-repo`) that declare the heading + route to `reflective-risk` | **Agree** |
+| FH | GLOSSARY playbook step 17 backfill (R85 preamble guards) + step 18 (composable Human Review guards) + governance sync | **Agree** |
+| FI | ROUTE holdout expansion | **Defer** |
+| FJ | Router / tenth skill / benchmark CI | **Reject** |
+
+### Round 86 verdict table
+
+| ID | Option | Verdict | Action |
+| --- | --- | --- | --- |
+| FG | Composable Human Review guards | **Agree** | parametrized tests via `prompt_eval_helpers.py`; align HR copy to `reflective-risk` |
+| FH | Playbook + docs | **Agree** | steps 17–18; panel round 86 sync |
+| FI | Holdout expansion | **Defer** | maintenance |
+| FJ | Router/tenth skill/benchmark CI | **Reject** | no change |
+
+**All roles agree.**
+
+## Implemented Changes (Round 86)
+
+- `plans/tests/prompt_eval_helpers.py`: exact `## Human Review` heading detection (avoids `## Human Review Required` false positives)
+- `plans/tests/test_{engineering,context,agent,domain,repo}_prompts_eval_harness.py`: Human Review preamble guards on prompts that declare the section
+- Prompt copy aligned to `reflective-risk` in `memory-consolidation.md`, `retro.md`, `small-context.md`, `creative-template.md`, `cursor-rules.md`, `PROJECT_KNOWLEDGE.template.md`; contract `## Human Review` added to `AGENTS.md`
+- `GLOSSARY.md`: playbook Rounds 1–86; step 17 (R85 backfill) + step 18 (composable Human Review guards)
+- `QUALITY_GATES_SUMMARY.md`: Human Review guard note; panel Rounds 1–86; 550+ pytest floor
+- `PROJECT_KNOWLEDGE.md`: Decision Index Round 86 entry
+- `README.md`, `reflective-prompt-library/README.md`, `test_readme_governance.py`: panel round 86 sync
+
+## Verification (Round 86)
+
+- `make all`: pytest + ROUTE-001/002/003 100%
+
+## Panel status (updated)
+
+**Resealed 2026-06-25** after **Round 86** (options FG–FJ). Composable prompts with `## Human Review` now have eval_harness preamble guards matching thinking-lens pattern (R81); full library contract parity closed (graph + primary surface + Human Review). Holdout expansion remains recurrence-gated maintenance.
+
+
diff --git a/reflective-prompt-library/plans/tests/prompt_eval_helpers.py b/reflective-prompt-library/plans/tests/prompt_eval_helpers.py
@@ -0,0 +1,18 @@
+"""Shared helpers for composable prompt eval_harness anti-drift tests."""
+
+import re
+from pathlib import Path
+
+HUMAN_REVIEW_HEADING = re.compile(r"^## Human Review\s*$", re.MULTILINE)
+
+
+def prompt_preamble(prompt_path: Path) -> str:
+    return prompt_path.read_text(encoding="utf-8").split("```", 1)[0]
+
+
+def has_human_review_preamble(prompt_path: Path) -> bool:
+    return bool(HUMAN_REVIEW_HEADING.search(prompt_preamble(prompt_path)))
+
+
+def prompts_with_human_review(prompts: tuple[Path, ...]) -> tuple[Path, ...]:
+    return tuple(p for p in prompts if has_human_review_preamble(p))
diff --git a/reflective-prompt-library/plans/tests/test_agent_prompts_eval_harness.py b/reflective-prompt-library/plans/tests/test_agent_prompts_eval_harness.py
@@ -6,8 +6,10 @@
 import pytest
 
 sys.path.insert(0, str(Path(__file__).parent.parent))
+sys.path.insert(0, str(Path(__file__).parent))
 
 from eval_harness import EvalHarness  # noqa: E402
+from prompt_eval_helpers import prompts_with_human_review  # noqa: E402
 
 AGENT_DIR = Path(__file__).parent.parent.parent / "04-agent"
 REPO_ROOT = str(Path(__file__).parent.parent.parent.parent)
@@ -21,6 +23,7 @@
 )
 
 AGENT_PROMPTS = tuple(sorted(AGENT_DIR.glob("*.md")))
+AGENT_PROMPTS_WITH_HUMAN_REVIEW = prompts_with_human_review(AGENT_PROMPTS)
 SUPPORTING_LENS_AGENT_PROMPTS = frozenset({"runtime-trust-boundary.md"})
 
 
@@ -46,7 +49,6 @@ def test_agent_prompt_meets_eval_harness_floor(prompt_path: Path, harness: EvalH
         f"{[(c['id'], c['result']) for c in result['checks']]}"
     )
 
-
 def test_agent_prompts_reference_workflow_skills():
     for prompt_path in AGENT_PROMPTS:
         text = prompt_path.read_text(encoding="utf-8")
@@ -64,6 +66,7 @@ def test_agent_prompts_cover_agent_workflow_surfaces():
     ):
         assert skill in text, f"04-agent should reference {skill}"
 
+
 def test_agent_prompts_have_workflow_surface_preamble_line():
     """04-agent prompts use Primary workflow surface(s) or Supporting lens (trust boundary)."""
     for prompt_path in AGENT_PROMPTS:
@@ -77,3 +80,14 @@ def test_agent_prompts_have_workflow_surface_preamble_line():
                 f"{prompt_path.name} Purpose should list Primary workflow surface(s)"
             )
 
+
+@pytest.mark.parametrize(
+    "prompt_path", AGENT_PROMPTS_WITH_HUMAN_REVIEW, ids=lambda p: p.name
+)
+def test_agent_prompt_has_human_review_section(prompt_path: Path):
+    """Prompts with Human Review declare escalation outside zh-TW templates."""
+    preamble = prompt_path.read_text(encoding="utf-8").split("```", 1)[0]
+    assert "## Human Review" in preamble, f"{prompt_path.name} missing Human Review preamble"
+    assert "reflective-risk" in preamble, (
+        f"{prompt_path.name} Human Review should route to reflective-risk"
+    )
diff --git a/reflective-prompt-library/plans/tests/test_context_prompts_eval_harness.py b/reflective-prompt-library/plans/tests/test_context_prompts_eval_harness.py
@@ -6,8 +6,10 @@
 import pytest
 
 sys.path.insert(0, str(Path(__file__).parent.parent))
+sys.path.insert(0, str(Path(__file__).parent))
 
 from eval_harness import EvalHarness  # noqa: E402
+from prompt_eval_helpers import prompts_with_human_review  # noqa: E402
 
 CONTEXT_DIR = Path(__file__).parent.parent.parent / "03-context"
 REPO_ROOT = str(Path(__file__).parent.parent.parent.parent)
@@ -21,6 +23,7 @@
 )
 
 CONTEXT_PROMPTS = tuple(sorted(CONTEXT_DIR.glob("*.md")))
+CONTEXT_PROMPTS_WITH_HUMAN_REVIEW = prompts_with_human_review(CONTEXT_PROMPTS)
 
 
 @pytest.fixture(scope="module")
@@ -45,7 +48,6 @@ def test_context_prompt_meets_eval_harness_floor(prompt_path: Path, harness: Eva
         f"{[(c['id'], c['result']) for c in result['checks']]}"
     )
 
-
 def test_context_prompts_reference_workflow_skills():
     for prompt_path in CONTEXT_PROMPTS:
         text = prompt_path.read_text(encoding="utf-8")
@@ -62,6 +64,7 @@ def test_context_prompts_cover_context_workflow_surfaces():
     ):
         assert skill in text, f"03-context should reference {skill}"
 
+
 def test_context_prompts_have_primary_workflow_surfaces_line():
     """All 03-context prompts declare Primary workflow surface(s) in Purpose preambles."""
     for prompt_path in CONTEXT_PROMPTS:
@@ -70,3 +73,14 @@ def test_context_prompts_have_primary_workflow_surfaces_line():
             f"{prompt_path.name} Purpose should list Primary workflow surface(s)"
         )
 
+
+@pytest.mark.parametrize(
+    "prompt_path", CONTEXT_PROMPTS_WITH_HUMAN_REVIEW, ids=lambda p: p.name
+)
+def test_context_prompt_has_human_review_section(prompt_path: Path):
+    """Prompts with Human Review declare escalation outside zh-TW templates."""
+    preamble = prompt_path.read_text(encoding="utf-8").split("```", 1)[0]
+    assert "## Human Review" in preamble, f"{prompt_path.name} missing Human Review preamble"
+    assert "reflective-risk" in preamble, (
+        f"{prompt_path.name} Human Review should route to reflective-risk"
+    )
diff --git a/reflective-prompt-library/plans/tests/test_domain_prompts_eval_harness.py b/reflective-prompt-library/plans/tests/test_domain_prompts_eval_harness.py
@@ -6,8 +6,10 @@
 import pytest
 
 sys.path.insert(0, str(Path(__file__).parent.parent))
+sys.path.insert(0, str(Path(__file__).parent))
 
 from eval_harness import EvalHarness  # noqa: E402
+from prompt_eval_helpers import prompts_with_human_review  # noqa: E402
 
 DOMAIN_DIR = Path(__file__).parent.parent.parent / "05-domain"
 REPO_ROOT = str(Path(__file__).parent.parent.parent.parent)
@@ -21,6 +23,7 @@
 )
 
 DOMAIN_PROMPTS = tuple(sorted(DOMAIN_DIR.glob("*.md")))
+DOMAIN_PROMPTS_WITH_HUMAN_REVIEW = prompts_with_human_review(DOMAIN_PROMPTS)
 
 
 @pytest.fixture(scope="module")
@@ -45,7 +48,6 @@ def test_domain_prompt_meets_eval_harness_floor(prompt_path: Path, harness: Eval
         f"{[(c['id'], c['result']) for c in result['checks']]}"
     )
 
-
 def test_domain_prompts_reference_workflow_skills():
     for prompt_path in DOMAIN_PROMPTS:
         text = prompt_path.read_text(encoding="utf-8")
@@ -64,11 +66,6 @@ def test_domain_prompts_cover_domain_workflow_surfaces():
         assert skill in text, f"05-domain should reference {skill}"
 
 
-def test_high_risk_prompt_has_human_review_section():
-    text = (DOMAIN_DIR / "high-risk.md").read_text(encoding="utf-8")
-    preamble = text.split("```", 1)[0]
-    assert "## Human Review" in preamble, "high-risk.md preamble should include Human Review"
-
 def test_domain_prompts_have_primary_workflow_surfaces_line():
     """All 05-domain prompts declare Primary workflow surface(s) in Purpose preambles."""
     for prompt_path in DOMAIN_PROMPTS:
@@ -77,3 +74,14 @@ def test_domain_prompts_have_primary_workflow_surfaces_line():
             f"{prompt_path.name} Purpose should list Primary workflow surface(s)"
         )
 
+
+@pytest.mark.parametrize(
+    "prompt_path", DOMAIN_PROMPTS_WITH_HUMAN_REVIEW, ids=lambda p: p.name
+)
+def test_domain_prompt_has_human_review_section(prompt_path: Path):
+    """Prompts with Human Review declare escalation outside zh-TW templates."""
+    preamble = prompt_path.read_text(encoding="utf-8").split("```", 1)[0]
+    assert "## Human Review" in preamble, f"{prompt_path.name} missing Human Review preamble"
+    assert "reflective-risk" in preamble, (
+        f"{prompt_path.name} Human Review should route to reflective-risk"
+    )
diff --git a/reflective-prompt-library/plans/tests/test_engineering_prompts_eval_harness.py b/reflective-prompt-library/plans/tests/test_engineering_prompts_eval_harness.py
diff --git a/reflective-prompt-library/plans/tests/test_glossary_structure.py b/reflective-prompt-library/plans/tests/test_glossary_structure.py
diff --git a/reflective-prompt-library/plans/tests/test_readme_governance.py b/reflective-prompt-library/plans/tests/test_readme_governance.py
diff --git a/reflective-prompt-library/plans/tests/test_repo_prompts_eval_harness.py b/reflective-prompt-library/plans/tests/test_repo_prompts_eval_harness.py
