feat: Phase L L2 reproducibility manifest schema -- exact versions, checksums, and seeds for third-party reproduction

Makefile

@@ -1,4 +1,4 @@
-.PHONY: help demo test lint ci clean bench-leakage bench-multi-negatives bench-baseline bench-hidden-active bench-cluster-split bench-expert-ablation bench-expert-ablation-500 bench-selectivity bench-feature-decomp bench-gate bench-easy-baseline bench-charge-matched bench-order-dependent bench-precision-at-k bench-per-family bench-simulation-ablation bench-simulation-ablation-within-amp simulation-gate bench-strategy-compare regenerate-all generate phase3 pilot validate-scoring validate-scoring-phase3 validate-scoring-strict external-pilot pilot-confident presynth-qc gold-standard diversity synthesis-order novelty-broad external-consensus questionnaire gate-check ip-report benchmark-card wave0-5-gate-check wave0-5-novelty-audit wave0-5-novelty-audit-v2 wave0-5-panel wave0-5-evidence wave0-5-fill-external wave0-5b-generate wave0-5b-filter recalibration-engine recalibration-engine-dry-run validate-policy-version generate-synthetic-lab-results bump-policy-version generate-review-packet failed-candidate-report safe-publication-filter classify-negative-informativeness validate-rejection-events negative-result-dashboard calibration-audit calibration-audit-example calibration-overfit-check result-quality-filter synthetic-result-policy-check calibration-decision-checklist calibration-rollback-plan simulation-registry validate-simulation-result-schema simulation-baseline-check adapter-gate-check simulation-provenance simulation-ensemble-check simulation-ci-report simulation-deprecation-check simulation-scope-check simulation-evidence-packet artifact-version candidate-manifest artifact-changelog integration-check adapter-author-check license-check artifact-compat-check contribution-check adoption-scorecard release-gate-check decision-log release-request-check coi-check rotation-plan-check security-report-check citation-check roadmap-sync-check advisory-review-check annual-review-check selection-rationale-check batch-priority-check pilot-package-check calibration-intake-check uncertainty-report-check preprint-bundle-check
+.PHONY: help demo test lint ci clean bench-leakage bench-multi-negatives bench-baseline bench-hidden-active bench-cluster-split bench-expert-ablation bench-expert-ablation-500 bench-selectivity bench-feature-decomp bench-gate bench-easy-baseline bench-charge-matched bench-order-dependent bench-precision-at-k bench-per-family bench-simulation-ablation bench-simulation-ablation-within-amp simulation-gate bench-strategy-compare regenerate-all generate phase3 pilot validate-scoring validate-scoring-phase3 validate-scoring-strict external-pilot pilot-confident presynth-qc gold-standard diversity synthesis-order novelty-broad external-consensus questionnaire gate-check ip-report benchmark-card wave0-5-gate-check wave0-5-novelty-audit wave0-5-novelty-audit-v2 wave0-5-panel wave0-5-evidence wave0-5-fill-external wave0-5b-generate wave0-5b-filter recalibration-engine recalibration-engine-dry-run validate-policy-version generate-synthetic-lab-results bump-policy-version generate-review-packet failed-candidate-report safe-publication-filter classify-negative-informativeness validate-rejection-events negative-result-dashboard calibration-audit calibration-audit-example calibration-overfit-check result-quality-filter synthetic-result-policy-check calibration-decision-checklist calibration-rollback-plan simulation-registry validate-simulation-result-schema simulation-baseline-check adapter-gate-check simulation-provenance simulation-ensemble-check simulation-ci-report simulation-deprecation-check simulation-scope-check simulation-evidence-packet artifact-version candidate-manifest artifact-changelog integration-check adapter-author-check license-check artifact-compat-check contribution-check adoption-scorecard release-gate-check decision-log release-request-check coi-check rotation-plan-check security-report-check citation-check roadmap-sync-check advisory-review-check annual-review-check selection-rationale-check batch-priority-check pilot-package-check calibration-intake-check uncertainty-report-check preprint-bundle-check reproducibility-manifest-check
 
 PYTHON := $(shell [ -f .venv/bin/python ] && echo .venv/bin/python || echo python3)
 PYTEST  := $(shell [ -f .venv/bin/pytest ] && echo .venv/bin/pytest || echo pytest)
@@ -915,3 +915,8 @@ preprint-bundle-check:
 	.venv/bin/openamp-foundry preprint-bundle-check \
 		--entry-json '{"bundle_id":"BND-001","batch_id":"BATCH-001","pipeline_version":"0.8.4","submission_date":"2026-07-09","title":"Computational nomination of AMP candidates from the OpenAMP Foundry","artifact_ids":["SEL-001","PRI-001","PKG-001","UQ-001","CAL-001"],"evidence_level":4,"preprint_doi":"10.1101/2026.07.09.000001","contact_email":"research@openamp.org","release_approved":true,"dry_lab_only":true}' \
 		--format text
+
+reproducibility-manifest-check:
+	.venv/bin/openamp-foundry reproducibility-manifest-check \
+		--entry-json '{"manifest_id":"RPM-001","batch_id":"BATCH-001","pipeline_version":"0.8.5","run_date":"2026-07-09","python_version":"3.11.9","package_checksums":{"openamp_foundry":"0.8.4","numpy":"1.26.4","pandas":"2.2.2","scikit-learn":"1.4.2","torch":"2.3.0"},"data_checksums":{"data/amp_sequences.fasta":"abc123def456"},"random_seeds":{"ensemble_sampling":42},"hardware_summary":"Apple M3 Pro, 36 GB RAM","reviewer":"alice","dry_lab_only":true}' \
+		--format text

docs/evidence/METRICS_CURRENT.md

@@ -5,7 +5,8 @@ Machine-readable snapshot: `outputs/metrics_snapshot.json` regenerated with `mak
 > **Purpose:** One authoritative table of current pipeline metrics. If any doc disagrees
 > with this file, this file wins. Updated whenever benchmark/benchmark config changes.
 >
-> **Last updated:** 2026-07-09 (Phase L L1 — preprint evidence bundle schema — v0.8.4)
+> **Last updated:** 2026-07-09 (Phase L L2 — reproducibility manifest schema — v0.8.5)
+> **New in v0.8.5:** Reproducibility manifest schema (L2 — Loop 125) — docs/evidence/REPRODUCIBILITY_MANIFEST_GUIDE.md with purpose, required field table (11 fields), package checksums format, data checksums format, warnings, validation workflow, honest-use boundary. src/openamp_foundry/evidence/reproducibility_manifest.py with ReproducibilityManifestEntry dataclass (11 fields: manifest_id, batch_id, pipeline_version, run_date, python_version, package_checksums, data_checksums, random_seeds, hardware_summary, reviewer, dry_lab_only), ReproducibilityManifestResult dataclass (6 fields: manifest_id, batch_id, package_count, data_file_count, passed, errors, warnings, dry_lab_only=True), MINIMUM_PACKAGES (3), MINIMUM_DATA_FILES (1), RECOMMENDED_PACKAGES (5), validate_reproducibility_manifest() (10 checks: RPM- prefix, non-empty batch_id/pipeline_version/python_version/hardware_summary/reviewer, YYYY-MM-DD run_date, package_checksums>=3, data_checksums>=1, dry_lab_only must be True; empty random_seeds warns, package_count<5 warns, hardware_summary contains 'unknown' warns), validate_reproducibility_manifest_dict() (10 required fields guard). CLI (openamp-foundry reproducibility-manifest-check) with --entry-json, --format text|json. make reproducibility-manifest-check target. Phase L L2: every pipeline run now has a machine-validated reproducibility record.
 > **New in v0.8.4:** Preprint evidence bundle schema (L1 — Loop 124) — docs/evidence/PREPRINT_BUNDLE_GUIDE.md with purpose, required field table (11 fields), minimum artifact count (3), evidence level guide, warnings, validation workflow, honest-use boundary. src/openamp_foundry/evidence/preprint_bundle.py with PreprintBundleEntry dataclass (11 fields: bundle_id, batch_id, pipeline_version, submission_date, title, artifact_ids, evidence_level, preprint_doi, contact_email, release_approved, dry_lab_only), PreprintBundleResult dataclass (5 fields: bundle_id, batch_id, artifact_count, passed, errors, warnings, dry_lab_only=True), MINIMUM_ARTIFACTS (3), RECOMMENDED_ARTIFACT_COUNT (5), MAX_TITLE_LENGTH (300), VALID_EVIDENCE_LEVELS (6: 1-6), validate_preprint_bundle() (10 checks: BND- prefix, non-empty batch_id/pipeline_version/contact_email, YYYY-MM-DD submission_date, non-empty title<=300 chars, artifact_ids>=3, valid evidence_level 1-6, release_approved must be True, dry_lab_only must be True; evidence_level<=2 warns, empty preprint_doi warns, artifact_count<5 warns), validate_preprint_bundle_dict() (10 required fields guard). CLI (openamp-foundry preprint-bundle-check) with --entry-json, --format text|json. make preprint-bundle-check target. Phase L L1: scientific preprints now have a machine-validated evidence bundle structure.
 > **New in v0.8.3:** Uncertainty quantification report schema (K5 — Loop 123) — docs/evidence/UNCERTAINTY_REPORT_GUIDE.md with purpose, required field table (12 fields), valid metric names (5), warnings, validation workflow, honest-use boundary. src/openamp_foundry/evidence/uncertainty_report.py with UncertaintyReportEntry dataclass (12 fields: report_id, batch_id, candidate_id, pipeline_version, metric_name, point_estimate, lower_bound, upper_bound, confidence_level, calibration_source, reviewer, dry_lab_only), UncertaintyReportResult dataclass (6 fields: report_id, candidate_id, interval_width, passed, errors, warnings, dry_lab_only=True), VALID_METRIC_NAMES (5: cytotoxicity_score, hemolysis_fraction, mic, selectivity_index, stability_score), WIDE_INTERVAL_THRESHOLD (10.0), validate_uncertainty_report() (11 checks: UQ- prefix, non-empty batch_id/candidate_id/pipeline_version/calibration_source/reviewer, valid metric_name, lower_bound<=point_estimate, upper_bound>=point_estimate, confidence_level 0.0-1.0, dry_lab_only must be True; wide interval warns, confidence<0.80 warns, confidence>0.99 warns), validate_uncertainty_report_dict() (11 required fields guard). CLI (openamp-foundry uncertainty-report-check) with --entry-json, --format text|json. make uncertainty-report-check target. Phase K K5: prediction intervals for external reviewers are now machine-validated.
 > **New in v0.8.2:** Post-experiment calibration intake schema (K4 — Loop 122) — docs/evidence/CALIBRATION_INTAKE_GUIDE.md with purpose, required field table (11 fields), note on dry_lab_only=False (real lab results), valid assay types (5), valid outcome values (4), warnings, validation workflow, honest-use boundary. src/openamp_foundry/evidence/calibration_intake.py with CalibrationIntakeEntry dataclass (11 fields: intake_id, batch_id, candidate_id, pipeline_version, assay_type, predicted_outcome, observed_outcome, predicted_confidence, intake_date, reviewer, dry_lab_only=False), CalibrationIntakeResult dataclass (6 fields: intake_id, candidate_id, prediction_correct, passed, errors, warnings, dry_lab_only=False), VALID_ASSAY_TYPES (5: cytotoxicity_assay, hemolysis_assay, membrane_disruption_assay, mic_assay, stability_assay), VALID_OUTCOME_VALUES (4: active, inactive, inconclusive, not_tested), validate_calibration_intake() (11 checks: CAL- prefix, non-empty batch_id/candidate_id/pipeline_version/reviewer, valid assay_type, valid predicted_outcome, valid observed_outcome, confidence 0.0-1.0, YYYY-MM-DD date, dry_lab_only must be False; high-confidence misprediction warns, inconclusive observed warns), validate_calibration_intake_dict() (10 required fields guard). CLI (openamp-foundry calibration-intake-check) with --entry-json, --format text|json. make calibration-intake-check target. Phase K K4: prediction-vs-outcome comparisons are now machine-validated with honest dry_lab_only=False enforcement.

docs/evidence/REPRODUCIBILITY_MANIFEST_GUIDE.md

@@ -0,0 +1,62 @@
+# Reproducibility Manifest Guide
+
+A *reproducibility manifest* captures the exact software versions, data checksums,
+and random seeds used in a pipeline run, enabling any third party to independently
+reproduce the results.
+
+## Purpose
+
+External researchers who receive a preprint evidence bundle need to be able to:
+- verify that the reported results can be reproduced from the stated inputs,
+- detect if intermediate data was silently modified,
+- reproduce the exact stochastic sampling used in ensemble methods.
+
+A reproducibility manifest gives them the information to do this.
+
+## Required fields
+
+| Field | Type | Constraint |
+|---|---|---|
+| `manifest_id` | str | Must start with `"RPM-"` |
+| `batch_id` | str | Non-empty |
+| `pipeline_version` | str | Non-empty |
+| `run_date` | str | ISO 8601 `YYYY-MM-DD` |
+| `python_version` | str | Non-empty (e.g. `"3.11.9"`) |
+| `package_checksums` | dict[str, str] | At least `MINIMUM_PACKAGES` (3) entries |
+| `data_checksums` | dict[str, str] | At least `MINIMUM_DATA_FILES` (1) entry |
+| `random_seeds` | dict[str, int] | May be empty; keys are step names |
+| `hardware_summary` | str | Non-empty (e.g. `"Apple M3 Pro, 36 GB RAM"`) |
+| `reviewer` | str | Non-empty |
+| `dry_lab_only` | bool | Must be `True` |
+
+## Package checksums
+
+`package_checksums` maps package name to version string or SHA-256 hash.
+Example: `{"openamp_foundry": "0.8.4", "numpy": "1.26.4"}`.
+
+## Data checksums
+
+`data_checksums` maps data file path (relative to repo root) to SHA-256 hex digest.
+Example: `{"data/amp_sequences.fasta": "abc123..."}`.
+
+## Warnings
+
+- `random_seeds` is empty — warning: no random seeds recorded; stochastic steps cannot be reproduced exactly.
+- `len(package_checksums) < 5` — warning: fewer than 5 packages recorded; consider adding key dependencies.
+- hardware_summary contains "unknown" — warning: hardware not identified; reproducibility may vary across platforms.
+
+## How to validate
+
+```bash
+make reproducibility-manifest-check
+# or
+openamp-foundry reproducibility-manifest-check \
+  --entry-json '{"manifest_id":"RPM-001","batch_id":"BATCH-001","pipeline_version":"0.8.5","run_date":"2026-07-09","python_version":"3.11.9","package_checksums":{"openamp_foundry":"0.8.4","numpy":"1.26.4","pandas":"2.2.2"},"data_checksums":{"data/amp_sequences.fasta":"abc123def456"},"random_seeds":{"ensemble_sampling":42},"hardware_summary":"Apple M3 Pro, 36 GB RAM","reviewer":"alice","dry_lab_only":true}' \
+  --format text
+```
+
+## Honest-use boundary
+
+Reproducibility manifests describe the computational environment, not the experimental
+conditions. They enable computational reproducibility only; wet-lab reproduction requires
+independent experimental protocols.

docs/research/ROADMAP.md

@@ -1,5 +1,8 @@
 # Roadmap
 
+### v0.8.5
+- Phase L L2: reproducibility manifest schema — captures exact software versions, data checksums, and random seeds for a pipeline run. ReproducibilityManifestEntry dataclass, validate_reproducibility_manifest(), CLI reproducibility-manifest-check.
+
 ### v0.8.4
 - Phase L L1: preprint evidence bundle schema — ties K-phase artifacts into a submission-ready record for scientific preprints. PreprintBundleEntry dataclass, validate_preprint_bundle(), CLI preprint-bundle-check.
 

src/openamp_foundry/cli/commands/reports.py

@@ -3022,3 +3022,37 @@ def _run_preprint_bundle_check(args: argparse.Namespace) -> int:
             print("  Preprint bundle validated. Evidence package is ready for submission review.")
 
     return 0 if result.passed else 3
+
+
+def _run_reproducibility_manifest_check(args: argparse.Namespace) -> int:
+    """Validate a reproducibility manifest entry."""
+    import json as _json
+    from openamp_foundry.evidence.reproducibility_manifest import validate_reproducibility_manifest_dict
+
+    try:
+        d = _json.loads(args.entry_json)
+    except _json.JSONDecodeError as exc:
+        print(_json.dumps({"status": "error", "error": f"Invalid JSON: {exc}"}))
+        return 2
+
+    if not isinstance(d, dict):
+        print(_json.dumps({"status": "error", "error": "--entry-json must be a JSON object"}))
+        return 2
+
+    result = validate_reproducibility_manifest_dict(d)
+    output_format = getattr(args, "format", "text")
+
+    if output_format == "json":
+        import dataclasses
+        print(_json.dumps(dataclasses.asdict(result), indent=2))
+    else:
+        status = "PASS" if result.passed else "FAIL"
+        print(f"Reproducibility manifest {result.manifest_id} (batch {result.batch_id}, {result.package_count} packages, {result.data_file_count} data files): {status}")
+        for e in result.errors:
+            print(f"  ERROR: {e}")
+        for w in result.warnings:
+            print(f"  WARN:  {w}")
+        if result.passed:
+            print("  Reproducibility manifest validated. Run is fully documented for third-party reproduction.")
+
+    return 0 if result.passed else 3

src/openamp_foundry/cli/main.py

@@ -15,7 +15,7 @@
 from openamp_foundry.cli.commands.reports import _run_security_report_check
 from openamp_foundry.cli.commands.reports import _run_citation_check
 from openamp_foundry.cli.commands.reports import _run_roadmap_sync_check
-from openamp_foundry.cli.commands.reports import _run_advisory_review_check, _run_annual_review_check, _run_selection_rationale_check, _run_batch_priority_check, _run_pilot_package_check, _run_calibration_intake_check, _run_uncertainty_report_check, _run_preprint_bundle_check
+from openamp_foundry.cli.commands.reports import _run_advisory_review_check, _run_annual_review_check, _run_selection_rationale_check, _run_batch_priority_check, _run_pilot_package_check, _run_calibration_intake_check, _run_uncertainty_report_check, _run_preprint_bundle_check, _run_reproducibility_manifest_check
 
 import argparse
 import json
@@ -2089,6 +2089,24 @@ def build_parser() -> argparse.ArgumentParser:
         help="Output format (default: text).",
     )
 
+    # ── Reproducibility manifest check (Phase L L2) ───────────────────
+    rmc = sub.add_parser(
+        "reproducibility-manifest-check",
+        help=(
+            "Validate a reproducibility manifest. Confirms software versions, "
+            "data checksums, and random seeds are documented. Dry-lab only."
+        ),
+    )
+    rmc.add_argument(
+        "--entry-json", type=str, required=True,
+        help="JSON string of a ReproducibilityManifestEntry dict (required).",
+    )
+    rmc.add_argument(
+        "--format", type=str, default="text",
+        choices=["text", "json"],
+        help="Output format (default: text).",
+    )
+
     # ── Selection rationale check (Phase K K1) ───────────────────────
     src2 = sub.add_parser(
         "selection-rationale-check",
@@ -2412,6 +2430,9 @@ def main(argv: list[str] | None = None) -> int:
     if args.command == "preprint-bundle-check":
         return _run_preprint_bundle_check(args)
 
+    if args.command == "reproducibility-manifest-check":
+        return _run_reproducibility_manifest_check(args)
+
     parser.error("unknown command")
     return 2