Add impact-by-impact native decomposition coverage gates (sprint 241)

2026-02-26 15:05:21 -07:00
parent e9dd06205e
commit 3c04216a54
12 changed files with 559 additions and 1 deletions
--- a/docs/constructive_editing_runtime_plan.md
+++ b/docs/constructive_editing_runtime_plan.md
@@ -150,7 +150,7 @@ Parallel planning tranche (active):
  - `logs/taskitem_runs/TEST_ONLY_spec_planning_baseline_20260226/deterministic_spec_readiness.json`

 Planning runtime controls (active):
- `sprint232_plan.md` to `sprint240_plan.md`
+- `sprint232_plan.md` to `sprint241_plan.md`
 - semantic bridge + intake augmentation + requirement injection + expansion gating are wired into:
  - `tools/mcp/run_sprint_taskitem_pipeline.sh`
 - current policy default is native-first semantic fallback:
@@ -169,3 +169,8 @@ Planning runtime controls (active):
  - `WSTONE_NATIVE_DECOMP_RETRY`
  - `WSTONE_NATIVE_DECOMP_TARGET_MIN_TASKS`
  - `native_decomposition_retry`
+- impact-specific native decomposition coverage can be enforced:
+  - `WSTONE_NATIVE_IMPACT_COVERAGE_GATE`
+  - `WSTONE_NATIVE_IMPACT_COVERAGE_ENFORCE`
+  - `WSTONE_NATIVE_IMPACT_COVERAGE_PROFILES`
+  - `native_impact_coverage`
--- a/docs/generator_readiness_gap_registry_2026-02-26.md
+++ b/docs/generator_readiness_gap_registry_2026-02-26.md
@@ -47,6 +47,7 @@ This is the canonical dated registry for "not production-ready" generator gaps.
 | GR-018 | Performance/security/rollout constrained refactor enforcement gap | `docs/gap_hunt_fullstack_multifile_2026-02-26.md`, `logs/taskitem_runs/challenging_fullstack_multifile_20260226_r5/results.jsonl` | `partial` | Hard checks now enforce migration rollback + data-loss policy, security deny-by-default, SLO p95 presence, and rollout staged+abort policy. Enforcement is contract-level; generator capability under these constraints is still weak in C++ AB path. | Constraint-aware generation follow-up |
 | GR-019 | Parity-blocked readiness load (gating without capability closure) | `logs/taskitem_runs/challenging_fullstack_multifile_20260226_r7/summary.json`, `logs/taskitem_runs/challenging_subset_prod_20260226_r7/summary.json`, `docs/sprint225_227_execution_tracker_2026-02-26.md` | `partial` | Sprint 225-227 reduced blocked parity load from `6` to `0` on both tracked hard catalogs while keeping unresolved divergence at `0`. This closes immediate safety debt for current corpora, but robustness is still contingent on pattern-driven repair classes. | Generalize repairs beyond queue-shaped transpile outputs |
 | GR-020 | Semantic fallback overuse masks weak native decomposition | `logs/taskitem_runs/TEST_ONLY_sprint236_semantic_fallback_audit_20260226/semantic_fallback_summary.json`, `logs/taskitem_runs/TEST_ONLY_sprint237_semantic_fallback_gate_fail_20260226/semantic_fallback_budget_gate.json`, `logs/taskitem_runs/TEST_ONLY_sprint238_native_gate_fail_20260226.json`, `logs/taskitem_runs/TEST_ONLY_sprint239_semantic_fallback_audit_20260226/semantic_fallback_summary.json`, `logs/taskitem_runs/TEST_ONLY_sprint240_retry_20260226_145631/00_summary.json`, `docs/sprint236_execution_tracker_2026-02-26.md`, `docs/sprint237_execution_tracker_2026-02-26.md`, `docs/sprint238_execution_tracker_2026-02-26.md`, `docs/sprint239_execution_tracker_2026-02-26.md`, `docs/sprint240_execution_tracker_2026-02-26.md` | `partial` | Sprint 236 added deterministic fallback-gap auditing with tool + constraint metadata. Sprint 237 added fallback-budget hard gate (`max-fallback-rate`) and produced expected fail/pass artifacts. Sprint 238 added native decomposition hard gate in pipeline (`min task count`, `min semantic signal count`) so weak native generation can be blocked before fallback masking. Sprint 239 added native reason enrichment and improved semantic signal density (`0 -> 6`). Sprint 240 added native decomposition retry with explicit minimum-task policy; on current sample retry attempted but did not improve depth (`2 -> 2`). | Native decomposition task-depth upgrade (increase native task granularity beyond 2) |
+| GR-021 | Impact-specific native decomposition coverage not enforced uniformly | `docs/native_decomposition_impact_list_2026-02-26.md`, `tools/mcp/profiles/native_decomposition_impact_profiles.json`, `logs/taskitem_runs/TEST_ONLY_sprint241_fullstack_impact_20260226_150416/00_summary.json`, `logs/taskitem_runs/TEST_ONLY_sprint241_native_impact_aggregate_20260226/native_impact_coverage_aggregate.json` | `partial` | Sprint 241 added one-by-one impact profiles and a pipeline-integrated native impact coverage gate (`native_impact_coverage`). Current fullstack sample activates 7 profiles and all fail, proving enforcement catches shallow coverage but generator closure is still pending. | Implement profile-specific native decomposition expansion to satisfy failing profile constraints |

 ## What Was Covered Today (Sprints 175-184)

--- a/docs/native_decomposition_impact_list_2026-02-26.md
+++ b/docs/native_decomposition_impact_list_2026-02-26.md
@@ -0,0 +1,68 @@
+# Native Decomposition Impact List - 2026-02-26
+
+Status: Active
+Purpose: Enumerate all known areas affected by shallow native decomposition and map each to enforceable tooling.
+
+## Impact List (One-by-One)
+
+1. API contract evolution
+- Problem: compatibility/versioning work collapses into generic tasks.
+- Tooling: `impact_api_contract_evolution` profile in `tools/mcp/profiles/native_decomposition_impact_profiles.json`.
+
+2. State migration and rollback safety
+- Problem: migration/backfill/rollback choreography omitted.
+- Tooling: `impact_state_migration_rollback` profile.
+
+3. Security policy propagation
+- Problem: cross-layer auth/deny-by-default checks missing.
+- Tooling: `impact_security_policy_propagation` profile.
+
+4. SLO/latency budget constraints
+- Problem: performance requirements missing from decomposition rationale.
+- Tooling: `impact_slo_latency_budget` profile.
+
+5. Rollout choreography
+- Problem: staged rollout/abort hooks not represented as tasks.
+- Tooling: `impact_rollout_choreography` profile.
+
+6. Multi-file transactional edits
+- Problem: partial cross-file updates pass as “done”.
+- Tooling: `impact_multifile_transactional_edit` profile.
+
+7. Concurrency/locking semantics
+- Problem: race/lock/deadlock work under-scoped.
+- Tooling: `impact_concurrency_locking` profile.
+
+8. Projection/environment constraints
+- Problem: embedded/mobile/serverless constraints ignored.
+- Tooling: `impact_projection_environment_constraints` profile.
+
+9. Polyglot boundary changes
+- Problem: SDK/client/server coupling not decomposed.
+- Tooling: `impact_polyglot_boundary_changes` profile.
+
+10. Observability/telemetry integrity
+- Problem: logging/metrics/tracing semantics omitted.
+- Tooling: `impact_observability_telemetry_integrity` profile.
+
+11. Data integrity and precision
+- Problem: financial/scientific precision risks hidden in generic tasks.
+- Tooling: `impact_data_integrity_precision` profile.
+
+12. Recovery/replay-critical flows
+- Problem: recovery/rollback/replay pathways not represented.
+- Tooling: `impact_disaster_recovery_replay` profile.
+
+## Enforcement Tooling
+
+- Checker:
+  - `tools/mcp/check_native_decomposition_impact_coverage.py`
+- Profiles:
+  - `tools/mcp/profiles/native_decomposition_impact_profiles.json`
+- Standalone gate wrapper:
+  - `tools/mcp/run_native_impact_coverage_gate.sh`
+- Pipeline integration toggles in `tools/mcp/run_sprint_taskitem_pipeline.sh`:
+  - `WSTONE_NATIVE_IMPACT_COVERAGE_GATE`
+  - `WSTONE_NATIVE_IMPACT_COVERAGE_ENFORCE`
+  - `WSTONE_NATIVE_IMPACT_COVERAGE_PROFILES`
+
--- a/docs/progress_log_2026-02-26.md
+++ b/docs/progress_log_2026-02-26.md
@@ -185,3 +185,22 @@
  - `logs/taskitem_runs/TEST_ONLY_sprint240_retry_20260226_145632/00_summary.json`
 - Current measured result:
  - retry attempted but not applied (`2 -> 2` tasks), confirming remaining native decomposition depth gap.
+
+## Sprint 241 Added (Same Day)
+
+- Added complete dated impact list for shallow native decomposition:
+  - `docs/native_decomposition_impact_list_2026-02-26.md`
+- Added one-by-one profile tooling:
+  - `tools/mcp/profiles/native_decomposition_impact_profiles.json`
+  - `tools/mcp/check_native_decomposition_impact_coverage.py`
+  - `tools/mcp/run_native_impact_coverage_gate.sh`
+  - `tools/mcp/analyze_native_impact_coverage.py`
+- Integrated impact-coverage gate into pipeline with optional enforcement:
+  - `WSTONE_NATIVE_IMPACT_COVERAGE_GATE`
+  - `WSTONE_NATIVE_IMPACT_COVERAGE_ENFORCE`
+  - `WSTONE_NATIVE_IMPACT_COVERAGE_PROFILES`
+- Dated artifacts:
+  - `logs/taskitem_runs/TEST_ONLY_sprint241_impact_gate_20260226_150403/00_summary.json`
+  - `logs/taskitem_runs/TEST_ONLY_sprint241_impact_gate_20260226_150404/06_native_impact_coverage.json`
+  - `logs/taskitem_runs/TEST_ONLY_sprint241_fullstack_impact_20260226_150416/00_summary.json`
+  - `logs/taskitem_runs/TEST_ONLY_sprint241_native_impact_aggregate_20260226/native_impact_coverage_aggregate.json`
--- a/docs/sprint241_execution_tracker_2026-02-26.md
+++ b/docs/sprint241_execution_tracker_2026-02-26.md
@@ -0,0 +1,55 @@
+# Sprint 241 Execution Tracker - 2026-02-26
+
+## Scope
+- `sprint241_plan.md`
+
+## Implemented
+
+Impact list (dated):
+- `docs/native_decomposition_impact_list_2026-02-26.md`
+
+Profile set (one-by-one):
+- `tools/mcp/profiles/native_decomposition_impact_profiles.json`
+- profiles cover 12 impact surfaces:
+  - API contract evolution
+  - state migration/rollback
+  - security propagation
+  - SLO/latency budget
+  - rollout choreography
+  - multi-file transactional edit
+  - concurrency/locking
+  - projection/environment constraints
+  - polyglot boundary changes
+  - observability/telemetry integrity
+  - data integrity/precision
+  - recovery/replay critical flows
+
+Checker + wrappers:
+- `tools/mcp/check_native_decomposition_impact_coverage.py`
+- `tools/mcp/run_native_impact_coverage_gate.sh`
+- `tools/mcp/analyze_native_impact_coverage.py`
+
+Pipeline integration in `tools/mcp/run_sprint_taskitem_pipeline.sh`:
+- `WSTONE_NATIVE_IMPACT_COVERAGE_GATE` (default `0`)
+- `WSTONE_NATIVE_IMPACT_COVERAGE_ENFORCE` (default `0`)
+- `WSTONE_NATIVE_IMPACT_COVERAGE_PROFILES` (profile path)
+- emits `06_native_impact_coverage.json`
+- summary includes `native_impact_coverage`
+
+## Baseline Validation Artifacts
+
+- soft-gate run:
+  - `logs/taskitem_runs/TEST_ONLY_sprint241_impact_gate_20260226_150403/00_summary.json`
+- hard-gate run (expected block):
+  - `logs/taskitem_runs/TEST_ONLY_sprint241_impact_gate_20260226_150404/06_native_impact_coverage.json`
+- multi-profile fullstack sample:
+  - `logs/taskitem_runs/TEST_ONLY_sprint241_fullstack_impact_20260226_150416/00_summary.json`
+- aggregate report:
+  - `logs/taskitem_runs/TEST_ONLY_sprint241_native_impact_aggregate_20260226/native_impact_coverage_aggregate.json`
+
+Observed signal:
+- Fullstack sample activated `7` profiles, all failing under current shallow native decomposition.
+
+## Explicit Completion Signal
+
+- Sprint 241: `DONE` (implemented + pipeline-integrated + enforced smoke verified)
--- a/editor/src/Sprint241IntegrationSummary.h
+++ b/editor/src/Sprint241IntegrationSummary.h
@@ -0,0 +1,6 @@
+#pragma once
+
+// Sprint 241 integration summary:
+// - Added comprehensive shallow-impact list and profile set.
+// - Added profile-driven native decomposition coverage checker and gate wrapper.
+// - Integrated optional impact coverage gating into sprint pipeline.
--- a/sprint241_plan.md
+++ b/sprint241_plan.md
@@ -0,0 +1,12 @@
+# Sprint 241 Plan: Impact-by-Impact Native Decomposition Coverage Gates
+
+## Goal
+Enumerate all major shallow-decomposition impact surfaces and enforce profile-specific native coverage checks for each.
+
+## Steps
+- Step 2248: Create complete dated impact list for shallow decomposition effects.
+- Step 2249: Define one-by-one impact profiles with trigger + required coverage constraints.
+- Step 2250: Implement native impact coverage checker over run artifacts.
+- Step 2251: Add standalone gate wrapper and aggregate analyzer.
+- Step 2252: Integrate optional impact coverage gate into sprint pipeline.
+- Step 2253: Add `Sprint241IntegrationSummary.h` and execution tracker.
--- a/tools/mcp/analyze_native_impact_coverage.py
+++ b/tools/mcp/analyze_native_impact_coverage.py
@@ -0,0 +1,61 @@
+#!/usr/bin/env python3
+import argparse
+import json
+from pathlib import Path
+from typing import Dict
+
+
+def load_json(path: Path) -> Dict:
+    with path.open("r", encoding="utf-8") as f:
+        return json.load(f)
+
+
+def write_json(path: Path, obj: Dict) -> None:
+    with path.open("w", encoding="utf-8") as f:
+        json.dump(obj, f, indent=2, sort_keys=True)
+        f.write("\n")
+
+
+def main() -> int:
+    p = argparse.ArgumentParser(description="Aggregate native impact coverage reports across runs.")
+    p.add_argument("--runs-root", default="logs/taskitem_runs")
+    p.add_argument("--include-glob", default="*")
+    p.add_argument("--out", required=True)
+    args = p.parse_args()
+
+    root = Path(args.runs_root)
+    profile_counts: Dict[str, Dict[str, int]] = {}
+    total_runs = 0
+    failed_runs = 0
+
+    for report_path in sorted(root.glob(f"{args.include_glob}/06_native_impact_coverage.json")):
+        try:
+            report = load_json(report_path)
+        except Exception:
+            continue
+        total_runs += 1
+        if report.get("status") == "fail":
+            failed_runs += 1
+        for c in report.get("checks", []):
+            pid = str(c.get("id", "unknown"))
+            row = profile_counts.setdefault(pid, {"active": 0, "passed": 0, "failed": 0})
+            row["active"] += 1
+            if c.get("passed", False):
+                row["passed"] += 1
+            else:
+                row["failed"] += 1
+
+    out = {
+        "status": "ok",
+        "total_runs": total_runs,
+        "failed_runs": failed_runs,
+        "fail_rate": round((failed_runs / total_runs), 4) if total_runs else 0.0,
+        "profile_counts": profile_counts,
+    }
+    write_json(Path(args.out), out)
+    print(json.dumps({"status": "ok", "total_runs": total_runs, "failed_runs": failed_runs, "out": args.out}, sort_keys=True))
+    return 0
+
+
+if __name__ == "__main__":
+    raise SystemExit(main())
--- a/tools/mcp/check_native_decomposition_impact_coverage.py
+++ b/tools/mcp/check_native_decomposition_impact_coverage.py
@@ -0,0 +1,166 @@
+#!/usr/bin/env python3
+import argparse
+import json
+from pathlib import Path
+from typing import Dict, List, Tuple
+
+
+def load_json(path: Path) -> Dict:
+    with path.open("r", encoding="utf-8") as f:
+        return json.load(f)
+
+
+def write_json(path: Path, obj: Dict) -> None:
+    with path.open("w", encoding="utf-8") as f:
+        json.dump(obj, f, indent=2, sort_keys=True)
+        f.write("\n")
+
+
+def load_native_tasks(run_dir: Path, summary: Dict) -> List[Dict]:
+    retry = summary.get("native_decomposition_retry") or {}
+    retry_applied = bool(retry.get("applied", False))
+    retry_path = run_dir / "02aa_generate_taskitems_retry.json"
+    base_path = run_dir / "02_generate_taskitems.json"
+    if retry_applied and retry_path.exists():
+        data = load_json(retry_path)
+        if isinstance(data, dict):
+            return list(data.get("tasks") or [])
+    if base_path.exists():
+        data = load_json(base_path)
+        if isinstance(data, dict):
+            return list(data.get("tasks") or [])
+    return []
+
+
+def load_spec_text(summary: Dict, repo_root: Path) -> str:
+    input_file = str(summary.get("input_file", "")).strip()
+    if not input_file:
+        return ""
+    p = Path(input_file)
+    if not p.is_absolute():
+        p = repo_root / input_file
+    try:
+        return p.read_text(encoding="utf-8", errors="ignore")
+    except Exception:
+        return ""
+
+
+def activate_profiles(profiles: List[Dict], spec_text: str) -> List[Dict]:
+    s = spec_text.lower()
+    active = []
+    for profile in profiles:
+        keys = [str(k).lower() for k in (profile.get("trigger_keywords") or [])]
+        if not keys:
+            continue
+        if any(k in s for k in keys):
+            active.append(profile)
+    return active
+
+
+def collect_task_unions(tasks: List[Dict]) -> Tuple[List[str], List[str], List[Dict]]:
+    reasons: List[str] = []
+    prereq_ops: List[str] = []
+    contracts: List[Dict] = []
+    for t in tasks:
+        rs = t.get("reasons") or []
+        po = t.get("prerequisiteOps") or []
+        ec = t.get("executionContract") or {}
+        for r in rs:
+            reasons.append(str(r).lower())
+        for op in po:
+            prereq_ops.append(str(op))
+        if isinstance(ec, dict):
+            contracts.append(ec)
+    return reasons, prereq_ops, contracts
+
+
+def check_profile(profile: Dict, tasks: List[Dict]) -> Dict:
+    reasons, prereq_ops, contracts = collect_task_unions(tasks)
+    missing: List[str] = []
+
+    min_tasks = int(profile.get("min_native_task_count", 0) or 0)
+    if len(tasks) < min_tasks:
+        missing.append(f"native_task_count<{min_tasks}")
+
+    req_reason_keys = [str(x).lower() for x in (profile.get("required_reason_keywords") or [])]
+    for key in req_reason_keys:
+        if not any(key in r for r in reasons):
+            missing.append(f"missing_reason_keyword:{key}")
+
+    req_ops = [str(x) for x in (profile.get("required_prerequisite_ops") or [])]
+    ops_set = set(prereq_ops)
+    for op in req_ops:
+        if op not in ops_set:
+            missing.append(f"missing_prerequisite_op:{op}")
+
+    req_contract = [str(x) for x in (profile.get("required_execution_contract") or [])]
+    for field in req_contract:
+        has_field_true = False
+        for ec in contracts:
+            if isinstance(ec.get(field), bool):
+                if ec.get(field) is True:
+                    has_field_true = True
+                    break
+            elif field in ec:
+                has_field_true = True
+                break
+        if not has_field_true:
+            missing.append(f"missing_execution_contract:{field}")
+
+    return {
+        "id": profile.get("id", "unknown"),
+        "description": profile.get("description", ""),
+        "passed": len(missing) == 0,
+        "native_task_count": len(tasks),
+        "missing": missing,
+    }
+
+
+def main() -> int:
+    parser = argparse.ArgumentParser(description="Check native decomposition coverage against impact profiles.")
+    parser.add_argument("--run-dir", required=True, help="Pipeline run directory containing 00_summary.json")
+    parser.add_argument("--profiles", default="tools/mcp/profiles/native_decomposition_impact_profiles.json", help="Profile JSON path")
+    parser.add_argument("--out", required=True, help="Output report path")
+    parser.add_argument("--enforce", action="store_true", help="Exit nonzero on failures")
+    args = parser.parse_args()
+
+    run_dir = Path(args.run_dir)
+    summary_path = run_dir / "00_summary.json"
+    if not summary_path.exists():
+        raise SystemExit(f"missing summary: {summary_path}")
+
+    summary = load_json(summary_path)
+    profiles_doc = load_json(Path(args.profiles))
+    profiles = list(profiles_doc.get("profiles") or [])
+    repo_root = Path(__file__).resolve().parents[2]
+    spec_text = load_spec_text(summary, repo_root)
+
+    active = activate_profiles(profiles, spec_text)
+    tasks = load_native_tasks(run_dir, summary)
+
+    checks = [check_profile(p, tasks) for p in active]
+    failing = [c for c in checks if not c.get("passed", False)]
+
+    result = {
+        "status": "ok" if not failing else "fail",
+        "run_dir": str(run_dir),
+        "active_profile_count": len(active),
+        "failing_profile_count": len(failing),
+        "active_profiles": [p.get("id", "unknown") for p in active],
+        "checks": checks,
+    }
+    write_json(Path(args.out), result)
+    print(json.dumps({
+        "status": result["status"],
+        "active_profile_count": result["active_profile_count"],
+        "failing_profile_count": result["failing_profile_count"],
+        "out": args.out,
+    }, sort_keys=True))
+
+    if args.enforce and failing:
+        return 12
+    return 0
+
+
+if __name__ == "__main__":
+    raise SystemExit(main())
--- a/tools/mcp/profiles/native_decomposition_impact_profiles.json
+++ b/tools/mcp/profiles/native_decomposition_impact_profiles.json
@@ -0,0 +1,113 @@
+{
+  "schema": "native_decomposition_impact_profiles_v1",
+  "profiles": [
+    {
+      "id": "impact_api_contract_evolution",
+      "description": "API/schema evolution with compatibility obligations.",
+      "trigger_keywords": ["api", "openapi", "schema", "contract", "compatibility", "version"],
+      "min_native_task_count": 5,
+      "required_reason_keywords": ["contract", "compatibility"],
+      "required_prerequisite_ops": ["whetstone_generate_taskitems", "whetstone_validate_taskitem", "whetstone_queue_ready"],
+      "required_execution_contract": ["deterministic", "rollbackRequired", "replayValidationRequired"]
+    },
+    {
+      "id": "impact_state_migration_rollback",
+      "description": "State/data migrations that require rollback and data safety choreography.",
+      "trigger_keywords": ["migration", "rollback", "backfill", "data loss", "compatibility window"],
+      "min_native_task_count": 6,
+      "required_reason_keywords": ["migration", "risk"],
+      "required_prerequisite_ops": ["whetstone_generate_taskitems", "whetstone_validate_taskitem", "whetstone_queue_ready"],
+      "required_execution_contract": ["deterministic", "rollbackRequired", "replayValidationRequired"]
+    },
+    {
+      "id": "impact_security_policy_propagation",
+      "description": "Cross-layer security policy propagation and deny-by-default handling.",
+      "trigger_keywords": ["security", "auth", "permission", "deny-by-default", "threat"],
+      "min_native_task_count": 5,
+      "required_reason_keywords": ["security", "risk"],
+      "required_prerequisite_ops": ["whetstone_validate_taskitem", "whetstone_queue_ready"],
+      "required_execution_contract": ["deterministic", "rollbackRequired"]
+    },
+    {
+      "id": "impact_slo_latency_budget",
+      "description": "Performance/SLO constrained changes.",
+      "trigger_keywords": ["latency", "slo", "p95", "throughput", "budget"],
+      "min_native_task_count": 4,
+      "required_reason_keywords": ["performance", "risk"],
+      "required_prerequisite_ops": ["whetstone_validate_taskitem"],
+      "required_execution_contract": ["deterministic", "replayValidationRequired"]
+    },
+    {
+      "id": "impact_rollout_choreography",
+      "description": "Feature flags, staged rollout, and abort choreography.",
+      "trigger_keywords": ["rollout", "stage", "feature flag", "canary", "abort"],
+      "min_native_task_count": 5,
+      "required_reason_keywords": ["rollout", "risk"],
+      "required_prerequisite_ops": ["whetstone_queue_ready", "whetstone_validate_taskitem"],
+      "required_execution_contract": ["deterministic", "rollbackRequired"]
+    },
+    {
+      "id": "impact_multifile_transactional_edit",
+      "description": "Cross-file transactional edits where partial updates are unsafe.",
+      "trigger_keywords": ["multi-file", "cross-file", "transaction", "refactor", "across files"],
+      "min_native_task_count": 6,
+      "required_reason_keywords": ["transaction", "contract"],
+      "required_prerequisite_ops": ["whetstone_queue_ready", "whetstone_validate_taskitem"],
+      "required_execution_contract": ["deterministic", "rollbackRequired", "replayValidationRequired"]
+    },
+    {
+      "id": "impact_concurrency_locking",
+      "description": "Threading, locking, races, and scheduling semantics.",
+      "trigger_keywords": ["concurrency", "thread", "mutex", "lock", "race", "deadlock"],
+      "min_native_task_count": 5,
+      "required_reason_keywords": ["concurrency", "risk"],
+      "required_prerequisite_ops": ["whetstone_validate_taskitem", "whetstone_queue_ready"],
+      "required_execution_contract": ["deterministic", "replayValidationRequired"]
+    },
+    {
+      "id": "impact_projection_environment_constraints",
+      "description": "Target projection/environment constraints (embedded/mobile/serverless/hardware limits).",
+      "trigger_keywords": ["arduino", "embedded", "mobile", "ios", "android", "serverless", "memory limit", "hardware"],
+      "min_native_task_count": 4,
+      "required_reason_keywords": ["capability", "constraint"],
+      "required_prerequisite_ops": ["whetstone_architect_intake", "whetstone_validate_taskitem"],
+      "required_execution_contract": ["deterministic"]
+    },
+    {
+      "id": "impact_polyglot_boundary_changes",
+      "description": "Cross-language boundary and API client/server regeneration work.",
+      "trigger_keywords": ["polyglot", "cross-language", "sdk", "client", "backend", "frontend"],
+      "min_native_task_count": 5,
+      "required_reason_keywords": ["compatibility", "contract"],
+      "required_prerequisite_ops": ["whetstone_generate_taskitems", "whetstone_validate_taskitem"],
+      "required_execution_contract": ["deterministic", "replayValidationRequired"]
+    },
+    {
+      "id": "impact_observability_telemetry_integrity",
+      "description": "Logging/metrics/tracing changes requiring consistent telemetry semantics.",
+      "trigger_keywords": ["observability", "telemetry", "metrics", "tracing", "logging"],
+      "min_native_task_count": 4,
+      "required_reason_keywords": ["semantic", "contract"],
+      "required_prerequisite_ops": ["whetstone_validate_taskitem"],
+      "required_execution_contract": ["deterministic"]
+    },
+    {
+      "id": "impact_data_integrity_precision",
+      "description": "Financial/scientific precision and data-integrity sensitive changes.",
+      "trigger_keywords": ["financial", "precision", "scientific", "numerical", "rounding"],
+      "min_native_task_count": 5,
+      "required_reason_keywords": ["risk", "contract"],
+      "required_prerequisite_ops": ["whetstone_validate_taskitem"],
+      "required_execution_contract": ["deterministic", "replayValidationRequired"]
+    },
+    {
+      "id": "impact_disaster_recovery_replay",
+      "description": "Recovery/replay/rollback critical workflows.",
+      "trigger_keywords": ["recovery", "replay", "rollback", "incident", "failover"],
+      "min_native_task_count": 5,
+      "required_reason_keywords": ["risk", "replay"],
+      "required_prerequisite_ops": ["whetstone_validate_taskitem", "whetstone_queue_ready"],
+      "required_execution_contract": ["deterministic", "rollbackRequired", "replayValidationRequired"]
+    }
+  ]
+}
--- a/tools/mcp/run_native_impact_coverage_gate.sh
+++ b/tools/mcp/run_native_impact_coverage_gate.sh
@@ -0,0 +1,28 @@
+#!/usr/bin/env bash
+set -euo pipefail
+
+ROOT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")/../.." && pwd)"
+RUN_DIR="${1:-}"
+if [[ -z "$RUN_DIR" ]]; then
+  echo "usage: $0 <run_dir> [out_json]" >&2
+  exit 2
+fi
+OUT_JSON="${2:-$RUN_DIR/06_native_impact_coverage.json}"
+PROFILES="${WSTONE_NATIVE_IMPACT_COVERAGE_PROFILES:-$ROOT_DIR/tools/mcp/profiles/native_decomposition_impact_profiles.json}"
+ENFORCE="${WSTONE_NATIVE_IMPACT_COVERAGE_ENFORCE:-1}"
+
+args=(
+  --run-dir "$RUN_DIR"
+  --profiles "$PROFILES"
+  --out "$OUT_JSON"
+)
+if [[ "$ENFORCE" == "1" ]]; then
+  args+=(--enforce)
+fi
+
+if python3 "$ROOT_DIR/tools/mcp/check_native_decomposition_impact_coverage.py" "${args[@]}" >/dev/null; then
+  jq -n --arg run_dir "$RUN_DIR" --argjson report "$(cat "$OUT_JSON")" '{status:"pass", run_dir:$run_dir, report:$report}'
+else
+  jq -n --arg run_dir "$RUN_DIR" --argjson report "$(cat "$OUT_JSON")" '{status:"fail", run_dir:$run_dir, report:$report}'
+  exit 12
+fi
--- a/tools/mcp/run_sprint_taskitem_pipeline.sh
+++ b/tools/mcp/run_sprint_taskitem_pipeline.sh
@@ -39,6 +39,9 @@ NATIVE_SEMANTIC_SIGNAL_MIN="${WSTONE_NATIVE_SEMANTIC_SIGNAL_MIN:-0}"
 NATIVE_REASON_ENRICHMENT="${WSTONE_NATIVE_REASON_ENRICHMENT:-1}"
 NATIVE_DECOMP_RETRY="${WSTONE_NATIVE_DECOMP_RETRY:-1}"
 NATIVE_DECOMP_TARGET_MIN_TASKS="${WSTONE_NATIVE_DECOMP_TARGET_MIN_TASKS:-5}"
+NATIVE_IMPACT_COVERAGE_GATE="${WSTONE_NATIVE_IMPACT_COVERAGE_GATE:-0}"
+NATIVE_IMPACT_COVERAGE_ENFORCE="${WSTONE_NATIVE_IMPACT_COVERAGE_ENFORCE:-0}"
+NATIVE_IMPACT_COVERAGE_PROFILES="${WSTONE_NATIVE_IMPACT_COVERAGE_PROFILES:-$ROOT_DIR/tools/mcp/profiles/native_decomposition_impact_profiles.json}"
 CAPABILITY_SIGNALS_JSON="${WSTONE_CAPABILITY_SIGNALS_JSON:-}"
 if [[ -z "$CAPABILITY_SIGNALS_JSON" ]]; then
  CAPABILITY_SIGNALS_JSON='{}'
@@ -93,6 +96,7 @@ SEMANTIC_GATE_JSON='{}'
 NATIVE_DECOMP_GATE_JSON='{}'
 NATIVE_REASON_ENRICHMENT_JSON='{}'
 NATIVE_DECOMP_RETRY_JSON='{}'
+NATIVE_IMPACT_COVERAGE_JSON='{}'
 if [[ "$SEMANTIC_PLANNING_BRIDGE" == "1" ]]; then
  python3 "$ROOT_DIR/tools/mcp/markdown_to_semantic_annotations.py" \
    --spec "$INPUT_FILE" \
@@ -571,6 +575,26 @@ SUMMARY_JSON="$(jq -nc \
    }
  }')"

+if [[ "$NATIVE_IMPACT_COVERAGE_GATE" == "1" ]]; then
+  # Emit a provisional summary so coverage checker can read run metadata.
+  printf '%s\n' "$SUMMARY_JSON" > "$OUT_DIR/00_summary.json"
+  if python3 "$ROOT_DIR/tools/mcp/check_native_decomposition_impact_coverage.py" \
+    --run-dir "$OUT_DIR" \
+    --profiles "$NATIVE_IMPACT_COVERAGE_PROFILES" \
+    --out "$OUT_DIR/06_native_impact_coverage.json" \
+    $([[ "$NATIVE_IMPACT_COVERAGE_ENFORCE" == "1" ]] && echo --enforce) >/dev/null; then
+    NATIVE_IMPACT_COVERAGE_JSON="$(cat "$OUT_DIR/06_native_impact_coverage.json")"
+  else
+    NATIVE_IMPACT_COVERAGE_JSON="$(cat "$OUT_DIR/06_native_impact_coverage.json")"
+    SUMMARY_JSON="$(printf '%s' "$SUMMARY_JSON" | jq --argjson nic "$NATIVE_IMPACT_COVERAGE_JSON" '.native_impact_coverage = $nic')"
+    printf '%s\n' "$SUMMARY_JSON" > "$OUT_DIR/00_summary.json"
+    echo "error: native impact coverage gate failed" >&2
+    echo "error: see $OUT_DIR/06_native_impact_coverage.json" >&2
+    exit 11
+  fi
+  SUMMARY_JSON="$(printf '%s' "$SUMMARY_JSON" | jq --argjson nic "$NATIVE_IMPACT_COVERAGE_JSON" '.native_impact_coverage = $nic')"
+fi
+
 if [[ "$CALIBRATE_AFTER_RUN" == "1" ]]; then
  CALIBRATION_OUT_DIR="$OUT_DIR/calibration"
  CALIBRATION_JSON="$(python3 "$ROOT_DIR/tools/mcp/analyze_taskitem_calibration.py" \