Добавить безопасное продолжение stage-loop

docs/orchestration/domain_scenario_loop_repo_adapter.md

@@ -96,6 +96,7 @@ python scripts/stage_agent_loop.py ingest-gui-run --manifest docs/orchestration/
 python scripts/stage_agent_loop.py prepare-repair --manifest docs/orchestration/<stage_loop>.json
 python scripts/stage_agent_loop.py run-repair --manifest docs/orchestration/<stage_loop>.json --dry-run
 python scripts/stage_agent_loop.py status --manifest docs/orchestration/<stage_loop>.json
+python scripts/stage_agent_loop.py continue --manifest docs/orchestration/<stage_loop>.json
 python scripts/stage_agent_loop.py summarize --manifest docs/orchestration/<stage_loop>.json
 ```
 
@@ -144,6 +145,8 @@ It stores the GUI review under `artifacts/domain_runs/stage_agent_loops/<stage_i
 
 Use `python scripts/stage_agent_loop.py status --manifest docs/orchestration/<stage_loop>.json` as the cheap read-only checkpoint before continuing a stage. It prints the current next action, closing gate, latest GUI run, latest repair coder status, and latest repair validation status without modifying artifacts.
 
+Use `python scripts/stage_agent_loop.py continue --manifest docs/orchestration/<stage_loop>.json` as the safe one-command continuation layer. It can prepare a repair iteration and materialize `run-repair --dry-run` automatically; it will not run the real coder pass unless `--execute-repair` is passed, and it waits for a `--run-id assistant-stage1-<id>` when the next required step is post-repair rerun/ingest validation.
+
 It also writes `stage_repair_handoff.md/json` next to the stage summary. That handoff is the preferred input for the next coder pass: it lists primary repair targets and sample user-facing failures without forcing the coder to reread the entire GUI conversation first.
 
 To prepare the next repair iteration from that handoff, run:

scripts/stage_agent_loop.py

@@ -954,6 +954,12 @@ def build_stage_repair_coder_command(
 
 
 def handle_run_repair(args: argparse.Namespace) -> int:
+    payload = run_stage_repair(args)
+    print(json.dumps(payload, ensure_ascii=False, indent=2))
+    return 0
+
+
+def run_stage_repair(args: argparse.Namespace) -> dict[str, Any]:
     stage_manifest_path = repo_path(args.manifest)
     stage_manifest = load_stage_manifest(stage_manifest_path)
     stage_dir = stage_dir_for(repo_path(args.output_root), stage_manifest["stage_id"])
@@ -990,8 +996,7 @@ def handle_run_repair(args: argparse.Namespace) -> int:
                 execution=payload,
             ),
         )
-        print(json.dumps(payload, ensure_ascii=False, indent=2))
-        return 0
+        return payload
 
     snapshots = dcl.snapshot_coder_candidate_files(repair_candidate_paths(plan))
     dcl.run_subprocess_command(
@@ -1019,11 +1024,16 @@ def handle_run_repair(args: argparse.Namespace) -> int:
             execution=payload,
         ),
     )
-    print(json.dumps(payload, ensure_ascii=False, indent=2))
-    return 0
+    return payload
 
 
 def handle_ingest_gui_run(args: argparse.Namespace) -> int:
+    summary = ingest_gui_run_review(args)
+    print(json.dumps(summary, ensure_ascii=False, indent=2))
+    return 0
+
+
+def ingest_gui_run_review(args: argparse.Namespace) -> dict[str, Any]:
     stage_manifest_path = repo_path(args.manifest)
     stage_manifest = load_stage_manifest(stage_manifest_path)
     stage_dir = stage_dir_for(repo_path(args.output_root), stage_manifest["stage_id"])
@@ -1053,36 +1063,49 @@ def handle_ingest_gui_run(args: argparse.Namespace) -> int:
     )
     save_stage_summary(stage_dir, summary)
     save_stage_repair_handoff(stage_dir, build_stage_repair_handoff(summary, review))
-    print(json.dumps(summary, ensure_ascii=False, indent=2))
-    return 0
+    return summary
 
 
 def handle_prepare_repair(args: argparse.Namespace) -> int:
-    stage_manifest_path = repo_path(args.manifest)
-    stage_manifest = load_stage_manifest(stage_manifest_path)
-    stage_dir = stage_dir_for(repo_path(args.output_root), stage_manifest["stage_id"])
-    handoff_path = repo_path(args.handoff) if args.handoff else stage_dir / "stage_repair_handoff.json"
-    handoff = load_json_object(handoff_path, "Stage repair handoff")
-    iteration_id = str(args.iteration_id or f"repair_{slugify(str(handoff.get('run_id') or 'gui_run'))}").strip()
+    payload = prepare_stage_repair_artifacts(
+        manifest_path=repo_path(args.manifest),
+        output_root=repo_path(args.output_root),
+        handoff_path=repo_path(args.handoff) if args.handoff else None,
+        iteration_id=args.iteration_id,
+    )
+    print(json.dumps(payload, ensure_ascii=False, indent=2))
+    return 0
+
+
+def prepare_stage_repair_artifacts(
+    *,
+    manifest_path: Path,
+    output_root: Path,
+    handoff_path: Path | None = None,
+    iteration_id: str | None = None,
+) -> dict[str, Any]:
+    stage_manifest = load_stage_manifest(manifest_path)
+    stage_dir = stage_dir_for(output_root, stage_manifest["stage_id"])
+    resolved_handoff_path = handoff_path if handoff_path is not None else stage_dir / "stage_repair_handoff.json"
+    handoff = load_json_object(resolved_handoff_path, "Stage repair handoff")
+    resolved_iteration_id = str(iteration_id or f"repair_{slugify(str(handoff.get('run_id') or 'gui_run'))}").strip()
     plan = build_stage_repair_iteration_plan(
         stage_manifest=stage_manifest,
         stage_dir=stage_dir,
         handoff=handoff,
-        iteration_id=iteration_id,
+        iteration_id=resolved_iteration_id,
     )
     iteration_dir = save_stage_repair_iteration(stage_dir, plan)
-    payload = {
+    return {
         "schema_version": "stage_gui_repair_prepare_result_v1",
         "stage_id": stage_manifest["stage_id"],
-        "iteration_id": iteration_id,
+        "iteration_id": resolved_iteration_id,
         "iteration_dir": repo_relative(iteration_dir),
         "repair_plan": repo_relative(iteration_dir / "repair_iteration_plan.json"),
         "repair_prompt": repo_relative(iteration_dir / "repair_prompt.md"),
         "repair_checklist": repo_relative(iteration_dir / "repair_checklist.md"),
         "candidate_files": plan.get("candidate_files") or [],
     }
-    print(json.dumps(payload, ensure_ascii=False, indent=2))
-    return 0
 
 
 def handle_plan(args: argparse.Namespace) -> int:
@@ -1105,6 +1128,7 @@ def handle_plan(args: argparse.Namespace) -> int:
 def build_stage_status(stage_manifest: dict[str, Any], stage_dir: Path) -> dict[str, Any]:
     summary_path = stage_dir / "stage_loop_summary.json"
     summary = load_json_object(summary_path, "Existing stage summary") if summary_path.exists() else {}
+    next_action = str(summary.get("next_action") or "run_stage_loop_or_ingest_gui_run")
     latest_gui_review = summary.get("latest_gui_review") if isinstance(summary.get("latest_gui_review"), dict) else {}
     latest_repair_execution = (
         summary.get("latest_repair_execution")
@@ -1132,8 +1156,8 @@ def build_stage_status(stage_manifest: dict[str, Any], stage_dir: Path) -> dict[
         "accepted_gate": summary.get("accepted_gate"),
         "loop_accepted_gate": summary.get("loop_accepted_gate"),
         "stage_closing_gate": stage_closing_gate or None,
-        "next_action": summary.get("next_action") or "run_stage_loop_or_ingest_gui_run",
-        "next_step_guidance": summary.get("next_step_guidance") or build_next_step_guidance("run_stage_loop_or_ingest_gui_run"),
+        "next_action": next_action,
+        "next_step_guidance": summary.get("next_step_guidance") or build_next_step_guidance(next_action),
         "latest_gui_run_id": latest_gui_review.get("run_id"),
         "latest_gui_business_status": latest_gui_review.get("overall_business_status"),
         "latest_repair_coder_status": latest_repair_execution.get("coder_status"),
@@ -1154,6 +1178,107 @@ def handle_status(args: argparse.Namespace) -> int:
     return 0
 
 
+def args_with(args: argparse.Namespace, **overrides: Any) -> argparse.Namespace:
+    values = vars(args).copy()
+    values.update(overrides)
+    return argparse.Namespace(**values)
+
+
+def handle_continue(args: argparse.Namespace) -> int:
+    stage_manifest_path = repo_path(args.manifest)
+    stage_manifest = load_stage_manifest(stage_manifest_path)
+    stage_dir = stage_dir_for(repo_path(args.output_root), stage_manifest["stage_id"])
+    stage_dir.mkdir(parents=True, exist_ok=True)
+    status_before = build_stage_status(stage_manifest, stage_dir)
+    next_action = str(status_before.get("next_action") or "")
+    payload: dict[str, Any] = {
+        "schema_version": "stage_agent_loop_continue_result_v1",
+        "stage_id": stage_manifest["stage_id"],
+        "stage_dir": repo_relative(stage_dir),
+        "observed_next_action": next_action,
+        "performed_action": "none",
+        "created_at": now_iso(),
+        "status_before": status_before,
+    }
+
+    if next_action in {"continue_repair_from_gui_review_p0", "continue_repair_from_gui_review_p1"}:
+        prepare_payload = prepare_stage_repair_artifacts(
+            manifest_path=stage_manifest_path,
+            output_root=repo_path(args.output_root),
+            handoff_path=repo_path(args.handoff) if args.handoff else None,
+            iteration_id=args.iteration_id,
+        )
+        repair_payload = run_stage_repair(
+            args_with(
+                args,
+                plan=str(repo_path(str(prepare_payload["repair_plan"]))),
+                dry_run=True,
+            )
+        )
+        payload.update(
+            {
+                "performed_action": "prepare_repair_and_run_repair_dry_run",
+                "prepare_repair": prepare_payload,
+                "run_repair": repair_payload,
+                "next_action": repair_payload.get("next_action"),
+            }
+        )
+    elif next_action == "execute_repair_without_dry_run_or_review_command":
+        if bool(args.execute_repair):
+            repair_payload = run_stage_repair(args_with(args, dry_run=False))
+            payload.update(
+                {
+                    "performed_action": "run_repair_execute",
+                    "run_repair": repair_payload,
+                    "next_action": repair_payload.get("next_action"),
+                }
+            )
+        else:
+            payload.update(
+                {
+                    "performed_action": "wait_for_explicit_execute_repair",
+                    "next_action": next_action,
+                    "suggested_command": (
+                        "python scripts/stage_agent_loop.py continue "
+                        "--manifest <stage_manifest.json> --execute-repair"
+                    ),
+                }
+            )
+    elif next_action == "rerun_same_stage_or_gui_and_ingest_result":
+        if getattr(args, "run_id", None):
+            ingest_summary = ingest_gui_run_review(args)
+            payload.update(
+                {
+                    "performed_action": "ingest_gui_run_for_repair_validation",
+                    "ingest_summary": ingest_summary,
+                    "next_action": ingest_summary.get("next_action"),
+                }
+            )
+        else:
+            payload.update(
+                {
+                    "performed_action": "wait_for_rerun_ingest",
+                    "next_action": next_action,
+                    "suggested_command": (
+                        "python scripts/stage_agent_loop.py continue "
+                        "--manifest <stage_manifest.json> --run-id assistant-stage1-<new_id>"
+                    ),
+                }
+            )
+    else:
+        payload.update(
+            {
+                "performed_action": "no_safe_automatic_step",
+                "next_action": next_action,
+            }
+        )
+
+    payload["status_after"] = build_stage_status(stage_manifest, stage_dir)
+    write_json(stage_dir / "stage_continue_result.json", payload)
+    print(json.dumps(payload, ensure_ascii=False, indent=2))
+    return 0
+
+
 def handle_summarize(args: argparse.Namespace) -> int:
     stage_manifest_path = repo_path(args.manifest)
     stage_manifest = load_stage_manifest(stage_manifest_path)
@@ -1260,6 +1385,23 @@ def build_parser() -> argparse.ArgumentParser:
     add_common_args(status_parser)
     status_parser.set_defaults(func=handle_status)
 
+    continue_parser = subparsers.add_parser(
+        "continue",
+        help="Execute the next safe stage-loop step derived from status.next_action.",
+    )
+    add_common_args(continue_parser)
+    continue_parser.add_argument("--handoff")
+    continue_parser.add_argument("--iteration-id")
+    continue_parser.add_argument("--plan")
+    continue_parser.add_argument("--coder-schema", default=str(DEFAULT_REPAIR_CODER_SCHEMA))
+    continue_parser.add_argument("--execute-repair", action="store_true")
+    continue_parser.add_argument("--run-id")
+    continue_parser.add_argument("--session-file")
+    continue_parser.add_argument("--sessions-dir", default=str(gui_review.DEFAULT_SESSIONS_DIR))
+    continue_parser.add_argument("--reports-dir", default=str(gui_review.DEFAULT_REPORTS_DIR))
+    continue_parser.add_argument("--review-output-dir")
+    continue_parser.set_defaults(func=handle_continue)
+
     ingest_parser = subparsers.add_parser(
         "ingest-gui-run",
         help="Attach an existing assistant-stage1 GUI run review to the stage loop summary.",

scripts/test_stage_agent_loop.py

@@ -592,6 +592,192 @@ class StageAgentLoopTests(unittest.TestCase):
         self.assertEqual(status["latest_repair_coder_status"], "patched")
         self.assertEqual(status["latest_validation_status"], "failed_p0")
 
+    def test_build_stage_status_derives_guidance_from_existing_next_action(self) -> None:
+        with tempfile.TemporaryDirectory() as tmp:
+            root = Path(tmp)
+            stage_dir = root / "stage_runs" / "agent_loop"
+            write_json(
+                stage_dir / "stage_loop_summary.json",
+                {
+                    "stage_id": "agent_loop",
+                    "next_action": "continue_repair_from_gui_review_p0",
+                },
+            )
+
+            status = stage_loop.build_stage_status(
+                {
+                    "stage_id": "agent_loop",
+                    "module_name": "Agent Loop",
+                    "title": "Agent Loop",
+                },
+                stage_dir,
+            )
+
+        self.assertEqual(status["next_action"], "continue_repair_from_gui_review_p0")
+        self.assertIn("prepare-repair", status["next_step_guidance"]["command_templates"][0])
+
+    def test_handle_continue_prepares_repair_and_dry_runs_command(self) -> None:
+        with tempfile.TemporaryDirectory() as tmp:
+            root = Path(tmp)
+            manifest_path = root / "stage.json"
+            output_root = root / "stage_runs"
+            stage_dir = output_root / "agent_loop"
+            schema_path = root / "coder.schema.json"
+            write_json(
+                manifest_path,
+                {
+                    "stage_id": "agent_loop",
+                    "module_name": "Agent Loop",
+                    "title": "Agent Loop",
+                    "pack_manifest": "docs/orchestration/demo_pack.json",
+                },
+            )
+            write_json(schema_path, {"type": "object", "additionalProperties": True})
+            write_json(
+                stage_dir / "stage_loop_summary.json",
+                {
+                    "stage_id": "agent_loop",
+                    "next_action": "continue_repair_from_gui_review_p0",
+                },
+            )
+            write_json(
+                stage_dir / "stage_repair_handoff.json",
+                {
+                    "run_id": "assistant-stage1-continue",
+                    "next_action": "continue_repair_from_gui_review_p0",
+                    "primary_repair_targets": [
+                        {
+                            "problem_layer": "answer_shape_mismatch",
+                            "issue_code": "business_direct_answer_missing",
+                            "severity": "P0",
+                            "occurrences": 1,
+                        }
+                    ],
+                    "sample_findings": [],
+                },
+            )
+
+            exit_code = stage_loop.handle_continue(
+                stage_args(
+                    manifest=str(manifest_path),
+                    output_root=str(output_root),
+                    handoff=None,
+                    iteration_id=None,
+                    plan=None,
+                    coder_schema=str(schema_path),
+                    execute_repair=False,
+                    run_id=None,
+                    session_file=None,
+                    sessions_dir=str(root / "sessions"),
+                    reports_dir=str(root / "reports"),
+                    review_output_dir=None,
+                )
+            )
+            result = json.loads((stage_dir / "stage_continue_result.json").read_text(encoding="utf-8"))
+            summary = json.loads((stage_dir / "stage_loop_summary.json").read_text(encoding="utf-8"))
+            command_exists = (
+                stage_dir
+                / "repair_iterations"
+                / "repair_assistant-stage1-continue"
+                / "repair_coder.command.txt"
+            ).exists()
+
+        self.assertEqual(exit_code, 0)
+        self.assertEqual(result["performed_action"], "prepare_repair_and_run_repair_dry_run")
+        self.assertEqual(result["next_action"], "execute_repair_without_dry_run_or_review_command")
+        self.assertEqual(summary["next_action"], "execute_repair_without_dry_run_or_review_command")
+        self.assertTrue(command_exists)
+
+    def test_handle_continue_waits_before_real_repair_without_flag(self) -> None:
+        with tempfile.TemporaryDirectory() as tmp:
+            root = Path(tmp)
+            manifest_path = root / "stage.json"
+            output_root = root / "stage_runs"
+            stage_dir = output_root / "agent_loop"
+            write_json(
+                manifest_path,
+                {
+                    "stage_id": "agent_loop",
+                    "module_name": "Agent Loop",
+                    "title": "Agent Loop",
+                    "pack_manifest": "docs/orchestration/demo_pack.json",
+                },
+            )
+            write_json(
+                stage_dir / "stage_loop_summary.json",
+                {
+                    "stage_id": "agent_loop",
+                    "next_action": "execute_repair_without_dry_run_or_review_command",
+                },
+            )
+
+            exit_code = stage_loop.handle_continue(
+                stage_args(
+                    manifest=str(manifest_path),
+                    output_root=str(output_root),
+                    handoff=None,
+                    iteration_id=None,
+                    plan=None,
+                    coder_schema=str(root / "coder.schema.json"),
+                    execute_repair=False,
+                    run_id=None,
+                    session_file=None,
+                    sessions_dir=str(root / "sessions"),
+                    reports_dir=str(root / "reports"),
+                    review_output_dir=None,
+                )
+            )
+            result = json.loads((stage_dir / "stage_continue_result.json").read_text(encoding="utf-8"))
+
+        self.assertEqual(exit_code, 0)
+        self.assertEqual(result["performed_action"], "wait_for_explicit_execute_repair")
+        self.assertEqual(result["next_action"], "execute_repair_without_dry_run_or_review_command")
+
+    def test_handle_continue_waits_for_rerun_ingest_without_run_id(self) -> None:
+        with tempfile.TemporaryDirectory() as tmp:
+            root = Path(tmp)
+            manifest_path = root / "stage.json"
+            output_root = root / "stage_runs"
+            stage_dir = output_root / "agent_loop"
+            write_json(
+                manifest_path,
+                {
+                    "stage_id": "agent_loop",
+                    "module_name": "Agent Loop",
+                    "title": "Agent Loop",
+                    "pack_manifest": "docs/orchestration/demo_pack.json",
+                },
+            )
+            write_json(
+                stage_dir / "stage_loop_summary.json",
+                {
+                    "stage_id": "agent_loop",
+                    "next_action": "rerun_same_stage_or_gui_and_ingest_result",
+                },
+            )
+
+            exit_code = stage_loop.handle_continue(
+                stage_args(
+                    manifest=str(manifest_path),
+                    output_root=str(output_root),
+                    handoff=None,
+                    iteration_id=None,
+                    plan=None,
+                    coder_schema=str(root / "coder.schema.json"),
+                    execute_repair=False,
+                    run_id=None,
+                    session_file=None,
+                    sessions_dir=str(root / "sessions"),
+                    reports_dir=str(root / "reports"),
+                    review_output_dir=None,
+                )
+            )
+            result = json.loads((stage_dir / "stage_continue_result.json").read_text(encoding="utf-8"))
+
+        self.assertEqual(exit_code, 0)
+        self.assertEqual(result["performed_action"], "wait_for_rerun_ingest")
+        self.assertEqual(result["next_action"], "rerun_same_stage_or_gui_and_ingest_result")
+
     def test_resolve_stage_repair_iteration_auto_prepares_from_handoff(self) -> None:
         with tempfile.TemporaryDirectory() as tmp:
             root = Path(tmp)