Add 5 documentation health metric guides

docs/evidence/METRIC_GUIDE_ARTIFACT_EXPLAINABILITY.md

@@ -0,0 +1,11 @@
+# Metric Guide: Artifact Explainability
+
+**Question:** Can a reviewer understand what a given artifact means and how it was produced?
+
+**What to track:** Whether each artifact type has a corresponding concept card or documentation section.
+
+**Target:** Every artifact type in ARTIFACT_MAP.md has a concept card.
+
+**What to measure:** Count of artifact types with concept cards divided by total artifact types.
+
+**What good looks like:** 100% coverage.

docs/evidence/METRIC_GUIDE_BENCHMARK_EXPLAINABILITY.md

@@ -0,0 +1,11 @@
+# Metric Guide: Benchmark Explainability
+
+**Question:** Can a reader understand what each benchmark measures and its limitations?
+
+**What to track:** Whether each benchmark in BENCHMARKING.md has a caveat section.
+
+**Target:** Every benchmark has at least one documented caveat.
+
+**What to measure:** Ratio of benchmarks with caveats to total benchmarks.
+
+**What good looks like:** A reader can find any benchmark and immediately see what it does NOT prove.

docs/evidence/METRIC_GUIDE_ONBOARDING_SUCCESS.md

@@ -0,0 +1,11 @@
+# Metric Guide: Onboarding Success
+
+**Question:** How quickly can a new contributor go from zero to running the demo?
+
+**What to track:** Time from `git clone` to `make demo` completing successfully.
+
+**Target:** < 15 minutes per QUICKSTART.md.
+
+**How to measure:** Run through QUICKSTART.md with a stopwatch.
+
+**What good looks like:** A contributor with Python installed can complete the full path in one sitting without asking for help.

docs/evidence/METRIC_GUIDE_REVIEW_READINESS.md

@@ -0,0 +1,16 @@
+# Metric Guide: Review Readiness
+
+**Question:** Is the project ready for external review?
+
+**What to track:** Whether all review artifacts exist and are up to date.
+
+**Artifacts to check:**
+- docs/review/EXPERT_REVIEW_PACK.md
+- docs/review/RELEASE_DRY_RUN_GUIDE.md
+- docs/review/PUBLICATION_PACK.md
+- docs/evidence/LIMITATIONS_OVERVIEW.md
+- .github/ISSUE_TEMPLATE/expert_review.yml
+
+**Target:** All 5 artifacts exist and pass claim check.
+
+**What good looks like:** A reviewer can be pointed to the review pack and have everything they need.

docs/evidence/METRIC_GUIDE_SAFETY_CLARITY.md

@@ -0,0 +1,11 @@
+# Metric Guide: Safety Clarity
+
+**Question:** Can a reviewer find the safety policy and understand what it permits?
+
+**What to track:** Time to find and understand the safety policy.
+
+**Target:** < 5 minutes.
+
+**What to measure:** A new reader can find SAFETY.md, identify allowed and forbidden uses, and state the dry-lab boundary within 5 minutes.
+
+**What good looks like:** Reviewer can quote the relevant policy section without assistance.

tests/test_metric_guides.py

@@ -0,0 +1,21 @@
+"""Tests for metric guides."""
+from pathlib import Path
+
+GUIDES = [
+    "METRIC_GUIDE_ONBOARDING_SUCCESS.md",
+    "METRIC_GUIDE_SAFETY_CLARITY.md",
+    "METRIC_GUIDE_ARTIFACT_EXPLAINABILITY.md",
+    "METRIC_GUIDE_BENCHMARK_EXPLAINABILITY.md",
+    "METRIC_GUIDE_REVIEW_READINESS.md",
+]
+
+
+def test_all_exist():
+    for g in GUIDES:
+        assert Path(f"docs/evidence/{g}").exists(), f"Missing: {g}"
+
+
+def test_each_has_question():
+    for g in GUIDES:
+        text = Path(f"docs/evidence/{g}").read_text()
+        assert "Question" in text or "question" in text