Add 10 tutorial templates and walkthroughs

docs/evidence/GUIDE_TUTORIAL_CLEANUP.md

@@ -0,0 +1,21 @@
+# Tutorial Cleanup Guidance
+
+Each tutorial should include cleanup instructions.
+
+## Format
+```markdown
+## Cleanup
+
+To remove the files created in this tutorial:
+
+```bash
+rm -rf outputs/demo_ranked.jsonl outputs/evidence/ outputs/demo_report.md
+```
+
+If you want to remove the entire project:
+
+```bash
+cd ..
+rm -rf OpenAMP-Foundry
+```
+```

docs/evidence/STANDARD_TUTORIAL_EXPECTED_FILES.md

@@ -0,0 +1,15 @@
+# Tutorial Expected Files Section Standard
+
+Each tutorial should include a section listing the files the reader should
+have after completing the tutorial.
+
+## Format
+```markdown
+## Expected Files
+
+After completing this tutorial, you should have:
+
+- `outputs/demo_ranked.jsonl` — ranked candidate list
+- `outputs/evidence/` — evidence certificates for selected candidates
+- `outputs/demo_report.md` — human-readable report
+```

docs/evidence/TEMPLATE_TUTORIAL_SAFETY_NOTE.md

@@ -0,0 +1,11 @@
+# Tutorial Safety Boundary Note Template
+
+Use this template for tutorials that touch safety-adjacent topics.
+
+```markdown
+> **Safety Note:** This tutorial demonstrates computational scoring only.
+> No biological activity has been demonstrated. The scores shown are
+> predictions, not proof of antimicrobial efficacy. Do not use these
+> results for clinical, safety, or release decisions without qualified
+> human review and wet-lab validation.
+```

docs/evidence/TEMPLATE_TUTORIAL_TROUBLESHOOTING.md

@@ -0,0 +1,16 @@
+# Tutorial Troubleshooting Appendix Template
+
+```markdown
+## Troubleshooting
+
+### `command not found: openamp-foundry`
+Make sure you've installed the package: `pip install -e ".[dev]"`
+
+### `ModuleNotFoundError: No module named 'openamp_foundry'`
+Make sure the virtual environment is activated and PYTHONPATH is set:
+`source .venv/bin/activate && export PYTHONPATH=src`
+
+### Tests fail
+Check your Python version: `python3 --version` (requires 3.11+)
+If tests still fail, open an issue.
+```

docs/evidence/TUTORIAL_INDEX_BY_LEVEL.md

@@ -0,0 +1,8 @@
+# Tutorial Index by Reader Level
+
+| Level | Reader | Recommended Tutorials |
+|-------|--------|---------------------|
+| Beginner | First-time contributor | QUICKSTART.md, first contribution decision tree |
+| Intermediate | Regular contributor | ARCHITECTURE.md, REVIEWER_CURRICULUM.md |
+| Advanced | Maintainer | Calibration audit walkthrough, release dry-run walkthrough |
+| Agent | AI agent | AGENTS.md, AGENT_TASKS.json, AGENT_CURRICULUM.md |

docs/evidence/TUTORIAL_PREREQS_CHECKLIST.md

@@ -0,0 +1,10 @@
+# Tutorial Prerequisites Checklist
+
+Before starting a tutorial, ensure:
+
+- [ ] Python 3.11+ is installed
+- [ ] `git` is installed
+- [ ] A virtual environment is created: `python3 -m venv .venv`
+- [ ] Dependencies are installed: `pip install -e ".[dev]"`
+- [ ] The tutorial's estimated time is accurate
+- [ ] All commands in the tutorial work from a clean checkout

docs/evidence/WALKTHROUGH_ARTIFACT_VALIDATION.md

@@ -0,0 +1,13 @@
+# Guided Walkthrough: Artifact Validation
+
+## Scenario
+You've received lab result JSON files from a partner. Validate them.
+
+## Steps
+1. Place the JSON files in a single directory.
+2. Run: `python scripts/validate_lab_data_return.py --results-dir <dir>`
+3. Check the output:
+   - "Passed: N" — all files passed schema validation
+   - "Failed: M" — some files have errors (check the error messages)
+4. If there are control failures, check the original assay data.
+5. If there are schema validation errors, ask the partner to fix and resubmit.

docs/evidence/WALKTHROUGH_BENCHMARK_REVIEW.md

@@ -0,0 +1,13 @@
+# Guided Walkthrough: Benchmark Report Review
+
+## Scenario
+You need to review a benchmark report to assess pipeline health.
+
+## Steps
+1. Run: `make bench-500`
+2. Check the AUROC value in `outputs/validate_scoring_500.json`.
+3. Compare with the expected value in METRICS_CURRENT.md.
+4. If AUROC dropped > 0.02, check for recent regressions.
+5. Run: `make bench-gate` — this checks all regression gates.
+6. Review the simulation ablation results in `outputs/simulation_ablation.json`.
+7. Check that simulation verdict remains NO_IMPROVEMENT.

docs/evidence/WALKTHROUGH_CALIBRATION_AUDIT.md

@@ -0,0 +1,13 @@
+# Guided Walkthrough: Calibration Audit Reports
+
+## Scenario
+You need to audit the calibration pipeline after receiving lab results.
+
+## Steps
+1. Place lab result JSONs in a directory.
+2. Run: `openamp-foundry calibration-intake --predictions <csv> --results-dir <dir> --panel-name wave1`
+3. Check the intake report for matched and orphan results.
+4. Run: `openamp-foundry recalibration-gate --intake <intake.json> --policy configs/recalibration_policy.yaml`
+5. Check the gate verdict: 0 = recalibration allowed, 3 = blocked.
+6. If blocked, review which conditions failed.
+7. Document the audit results in the decision log.

docs/evidence/WALKTHROUGH_REVIEW_PACKETS.md

@@ -0,0 +1,13 @@
+# Guided Walkthrough: Review Packets
+
+## Scenario
+You need to prepare a review packet for an external expert.
+
+## Steps
+1. Collect the evidence certificates for selected candidates.
+2. Run: `make lab-batch-pack` to create a zip with all artifacts.
+3. Include LIMITATIONS_OVERVIEW.md for honest caveats.
+4. Include METRICS_CURRENT.md for benchmark context.
+5. Include the specific claim or question being reviewed.
+6. Run `python scripts/safety/check_claims.py` on all documents.
+7. Send the packet with a clear description of what feedback is needed.

tests/test_tutorials_batch.py

@@ -0,0 +1,20 @@
+"""Tests for tutorials and walkthroughs batch."""
+from pathlib import Path
+
+FILES = [
+    "TUTORIAL_INDEX_BY_LEVEL.md",
+    "TUTORIAL_PREREQS_CHECKLIST.md",
+    "STANDARD_TUTORIAL_EXPECTED_FILES.md",
+    "GUIDE_TUTORIAL_CLEANUP.md",
+    "TEMPLATE_TUTORIAL_SAFETY_NOTE.md",
+    "TEMPLATE_TUTORIAL_TROUBLESHOOTING.md",
+    "WALKTHROUGH_ARTIFACT_VALIDATION.md",
+    "WALKTHROUGH_BENCHMARK_REVIEW.md",
+    "WALKTHROUGH_REVIEW_PACKETS.md",
+    "WALKTHROUGH_CALIBRATION_AUDIT.md",
+]
+
+
+def test_all_exist():
+    for f in FILES:
+        assert Path("docs/evidence", f).exists(), f"Missing: {f}"