devagent-physical-engine 1.4.0__tar.gz → 1.4.1__tar.gz

{devagent_physical_engine-1.4.0/src/devagent_physical_engine.egg-info → devagent_physical_engine-1.4.1}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: devagent-physical-engine
-Version: 1.4.0
+Version: 1.4.1
 Summary: Production-grade verification-first robotics evidence runtime with one-command customer intake, selective verification execution, deterministic Twin lineage, measured physical verification, replay inspection, regression, and FAT evidence.
 Author: Tom Ha
 License: All Rights Reserved
@@ -80,7 +80,7 @@ DevAgent is designed to answer: **What can be proven about this cell, what faile
 Python 3.11+ is required.
 
 ```bash
-python -m pip install --upgrade devagent-physical-engine==1.4.0
+python -m pip install --upgrade devagent-physical-engine==1.4.1
 ```
 
 Verify the package:
@@ -92,7 +92,7 @@ python -c "import devagent_physical_engine as d; print(d.__version__)"
 Expected:
 
 ```text
-1.4.0
+1.4.1
 ```
 
 Run the normal customer workflow:
@@ -720,9 +720,9 @@ Existing certified controllers execute.
 
 # Project status
 
-**v1.4.0 — Production/Stable software workflow**
+**v1.4.1 — Production/Stable software workflow**
 
-v1.4.0 adds selective verification execution while preserving the existing fail-closed authority model.
+v1.4.1 preserves v1.4 selective verification and fixes repeated full `devagent verify` runs by automatically binding exact previous-campaign → current-campaign regression evidence before FAT generation. The original FAT regression scope validation remains authoritative and fail-closed.
 
 Production-oriented software capabilities include:
 
@@ -745,7 +745,7 @@ change impact + regression
 provider-neutral optional AI front end
 ```
 
-Quality gates include Python 3.11/3.12/3.13 regression, package clean-install, dependency audit, global branch coverage, One-Command coverage, **Selective Verification coverage**, Production Authority coverage, and Evidence Trust/Replay coverage.
+Quality gates include Python 3.11/3.12/3.13 regression, package clean-install, dependency audit, global branch coverage, One-Command coverage, repeat-run regression qualification, **Selective Verification coverage**, Production Authority coverage, and Evidence Trust/Replay coverage.
 
 Production/Stable describes the bounded software workflow. It does not claim arbitrary customer cells are physically qualified, site-qualified, or safe for autonomous real execution.
 

{devagent_physical_engine-1.4.0 → devagent_physical_engine-1.4.1}/README.md

@@ -39,7 +39,7 @@ DevAgent is designed to answer: **What can be proven about this cell, what faile
 Python 3.11+ is required.
 
 ```bash
-python -m pip install --upgrade devagent-physical-engine==1.4.0
+python -m pip install --upgrade devagent-physical-engine==1.4.1
 ```
 
 Verify the package:
@@ -51,7 +51,7 @@ python -c "import devagent_physical_engine as d; print(d.__version__)"
 Expected:
 
 ```text
-1.4.0
+1.4.1
 ```
 
 Run the normal customer workflow:
@@ -679,9 +679,9 @@ Existing certified controllers execute.
 
 # Project status
 
-**v1.4.0 — Production/Stable software workflow**
+**v1.4.1 — Production/Stable software workflow**
 
-v1.4.0 adds selective verification execution while preserving the existing fail-closed authority model.
+v1.4.1 preserves v1.4 selective verification and fixes repeated full `devagent verify` runs by automatically binding exact previous-campaign → current-campaign regression evidence before FAT generation. The original FAT regression scope validation remains authoritative and fail-closed.
 
 Production-oriented software capabilities include:
 
@@ -704,7 +704,7 @@ change impact + regression
 provider-neutral optional AI front end
 ```
 
-Quality gates include Python 3.11/3.12/3.13 regression, package clean-install, dependency audit, global branch coverage, One-Command coverage, **Selective Verification coverage**, Production Authority coverage, and Evidence Trust/Replay coverage.
+Quality gates include Python 3.11/3.12/3.13 regression, package clean-install, dependency audit, global branch coverage, One-Command coverage, repeat-run regression qualification, **Selective Verification coverage**, Production Authority coverage, and Evidence Trust/Replay coverage.
 
 Production/Stable describes the bounded software workflow. It does not claim arbitrary customer cells are physically qualified, site-qualified, or safe for autonomous real execution.
 

{devagent_physical_engine-1.4.0 → devagent_physical_engine-1.4.1}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "devagent-physical-engine"
-version = "1.4.0"
+version = "1.4.1"
 description = "Production-grade verification-first robotics evidence runtime with one-command customer intake, selective verification execution, deterministic Twin lineage, measured physical verification, replay inspection, regression, and FAT evidence."
 readme = { file = "README.md", content-type = "text/markdown" }
 requires-python = ">=3.11"
@@ -47,7 +47,7 @@ Repository = "https://github.com/tomha85/devagent-physical-engine"
 Issues = "https://github.com/tomha85/devagent-physical-engine/issues"
 
 [project.scripts]
-devagent = "devagent_physical_engine.user_cli:main"
+devagent = "devagent_physical_engine.one_command_runtime:main"
 devagent-physical = "devagent_physical_engine.cli:main"
 devagent-physical-ai = "devagent_physical_engine.ai_cli:main"
 devagent-commercial = "devagent_physical_engine.commercial_cli:main"

{devagent_physical_engine-1.4.0 → devagent_physical_engine-1.4.1}/src/devagent_physical_engine/__init__.py

@@ -50,6 +50,7 @@ from .production_grade import (
     VerificationCase,
     load_verification_spec,
 )
+from .one_command_runtime import install_one_command_runtime
 from .project import (
     EvidenceArtifactRecord,
     ProjectError,
@@ -149,4 +150,6 @@ __all__ = [
     "VerificationTestMappingRecorder",
 ]
 
-__version__ = "1.4.0"
+install_one_command_runtime()
+
+__version__ = "1.4.1"

devagent_physical_engine-1.4.1/src/devagent_physical_engine/one_command_runtime.py

@@ -0,0 +1,114 @@
+"""Runtime hardening for the one-command customer workflow.
+
+This module keeps the production FAT builder fail-closed while making repeated
+``devagent verify`` runs lineage-complete. A second or later full campaign must
+carry an exact previous-campaign -> current-campaign regression artifact before
+FAT generation. The runtime creates that artifact automatically, then delegates
+back to :class:`ProductionFATReportBuilder` so the existing scope checks remain
+authoritative.
+"""
+
+from __future__ import annotations
+
+import json
+import sys
+from typing import Sequence
+
+from .commercial import CommercialError
+from .production_grade import ProductionFATReportBuilder, ProductionRegressionAnalyzer
+
+
+class OneCommandFATReportBuilder(ProductionFATReportBuilder):
+    """FAT builder adapter that supplies exact repeat-run regression lineage."""
+
+    def build(
+        self,
+        *,
+        project_id: str,
+        campaign_artifact_id: str,
+        regression_artifact_id: str | None = None,
+    ):
+        resolved_regression_id = regression_artifact_id
+        campaigns = self.repository.list_artifacts(
+            project_id,
+            artifact_type="commercial_campaign",
+        )
+        campaign_ids = [item.artifact_id for item in campaigns]
+
+        if campaign_artifact_id in campaign_ids and resolved_regression_id is None:
+            position = campaign_ids.index(campaign_artifact_id)
+            if position > 0:
+                baseline = campaigns[position - 1]
+                current = campaigns[position]
+
+                # Reuse an already-recorded exact lineage if a caller retries FAT
+                # generation. Otherwise create the immutable regression artifact.
+                regressions = self.repository.list_artifacts(
+                    project_id,
+                    artifact_type="commercial_regression",
+                )
+                for regression in reversed(regressions):
+                    if (
+                        regression.payload.get("baseline_campaign") == baseline.artifact_id
+                        and regression.payload.get("current_campaign") == current.artifact_id
+                    ):
+                        resolved_regression_id = regression.artifact_id
+                        break
+
+                if resolved_regression_id is None:
+                    regression = ProductionRegressionAnalyzer(self.repository).compare(
+                        project_id=project_id,
+                        baseline_artifact_id=baseline.artifact_id,
+                        current_artifact_id=current.artifact_id,
+                    )
+                    resolved_regression_id = regression.artifact_id
+
+        # Do not bypass or weaken ProductionFATReportBuilder._scoped_regression.
+        # It still validates the exact previous/current campaign relationship.
+        return super().build(
+            project_id=project_id,
+            campaign_artifact_id=campaign_artifact_id,
+            regression_artifact_id=resolved_regression_id,
+        )
+
+
+def install_one_command_runtime() -> None:
+    """Install the repeat-run-aware FAT adapter into Simple Mode idempotently."""
+
+    from . import one_command as one_command_module
+
+    if one_command_module.ProductionFATReportBuilder is not OneCommandFATReportBuilder:
+        one_command_module.ProductionFATReportBuilder = OneCommandFATReportBuilder
+
+
+def _bounded_failure(code: str, *, json_only: bool) -> int:
+    if not json_only:
+        print(f"[DevAgent] FAILED ERROR {code}", file=sys.stderr, flush=True)
+    print(
+        json.dumps(
+            {
+                "passed": False,
+                "failure_code": code,
+                "real_execution_allowed": False,
+            },
+            indent=2,
+            sort_keys=True,
+        )
+    )
+    return 10
+
+
+def main(argv: Sequence[str] | None = None) -> int:
+    """Public ``devagent`` entry point with bounded commercial-error handling."""
+
+    install_one_command_runtime()
+    from . import user_cli
+
+    effective_argv = list(sys.argv[1:] if argv is None else argv)
+    try:
+        return user_cli.main(effective_argv)
+    except CommercialError as exc:
+        # Expected engineering/business-state failures must not leak a Python
+        # traceback to a customer-facing CLI. Unexpected exception classes are
+        # intentionally not swallowed here.
+        return _bounded_failure(str(exc), json_only="--json" in effective_argv)

{devagent_physical_engine-1.4.0 → devagent_physical_engine-1.4.1/src/devagent_physical_engine.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: devagent-physical-engine
-Version: 1.4.0
+Version: 1.4.1
 Summary: Production-grade verification-first robotics evidence runtime with one-command customer intake, selective verification execution, deterministic Twin lineage, measured physical verification, replay inspection, regression, and FAT evidence.
 Author: Tom Ha
 License: All Rights Reserved
@@ -80,7 +80,7 @@ DevAgent is designed to answer: **What can be proven about this cell, what faile
 Python 3.11+ is required.
 
 ```bash
-python -m pip install --upgrade devagent-physical-engine==1.4.0
+python -m pip install --upgrade devagent-physical-engine==1.4.1
 ```
 
 Verify the package:
@@ -92,7 +92,7 @@ python -c "import devagent_physical_engine as d; print(d.__version__)"
 Expected:
 
 ```text
-1.4.0
+1.4.1
 ```
 
 Run the normal customer workflow:
@@ -720,9 +720,9 @@ Existing certified controllers execute.
 
 # Project status
 
-**v1.4.0 — Production/Stable software workflow**
+**v1.4.1 — Production/Stable software workflow**
 
-v1.4.0 adds selective verification execution while preserving the existing fail-closed authority model.
+v1.4.1 preserves v1.4 selective verification and fixes repeated full `devagent verify` runs by automatically binding exact previous-campaign → current-campaign regression evidence before FAT generation. The original FAT regression scope validation remains authoritative and fail-closed.
 
 Production-oriented software capabilities include:
 
@@ -745,7 +745,7 @@ change impact + regression
 provider-neutral optional AI front end
 ```
 
-Quality gates include Python 3.11/3.12/3.13 regression, package clean-install, dependency audit, global branch coverage, One-Command coverage, **Selective Verification coverage**, Production Authority coverage, and Evidence Trust/Replay coverage.
+Quality gates include Python 3.11/3.12/3.13 regression, package clean-install, dependency audit, global branch coverage, One-Command coverage, repeat-run regression qualification, **Selective Verification coverage**, Production Authority coverage, and Evidence Trust/Replay coverage.
 
 Production/Stable describes the bounded software workflow. It does not claim arbitrary customer cells are physically qualified, site-qualified, or safe for autonomous real execution.
 

{devagent_physical_engine-1.4.0 → devagent_physical_engine-1.4.1}/src/devagent_physical_engine.egg-info/SOURCES.txt

@@ -19,6 +19,7 @@ src/devagent_physical_engine/evidence_trust.py
 src/devagent_physical_engine/execution.py
 src/devagent_physical_engine/models.py
 src/devagent_physical_engine/one_command.py
+src/devagent_physical_engine/one_command_runtime.py
 src/devagent_physical_engine/operating_envelope.py
 src/devagent_physical_engine/physical_campaign.py
 src/devagent_physical_engine/physical_evidence.py
@@ -148,6 +149,7 @@ tests/test_moveit_planning_diagnostics_v0125.py
 tests/test_moveit_scene_v0123.py
 tests/test_natural_language_agent_v07.py
 tests/test_one_command_hardening_v13.py
+tests/test_one_command_regression_v141.py
 tests/test_one_command_v13.py
 tests/test_openai_strict_schema_v0113.py
 tests/test_optimization_v05.py

{devagent_physical_engine-1.4.0 → devagent_physical_engine-1.4.1}/src/devagent_physical_engine.egg-info/entry_points.txt

@@ -1,5 +1,5 @@
 [console_scripts]
-devagent = devagent_physical_engine.user_cli:main
+devagent = devagent_physical_engine.one_command_runtime:main
 devagent-commercial = devagent_physical_engine.commercial_cli:main
 devagent-physical = devagent_physical_engine.cli:main
 devagent-physical-ai = devagent_physical_engine.ai_cli:main

devagent_physical_engine-1.4.1/tests/test_one_command_regression_v141.py

@@ -0,0 +1,141 @@
+from __future__ import annotations
+
+from contextlib import redirect_stderr, redirect_stdout
+import io
+import json
+from pathlib import Path
+import tempfile
+import unittest
+from unittest.mock import patch
+
+import yaml
+
+from devagent_physical_engine.commercial import CommercialError
+from devagent_physical_engine.commercial_repository import CommercialProjectRepository
+from devagent_physical_engine.one_command import OneCommandVerifier
+from devagent_physical_engine.one_command_runtime import main as runtime_main
+
+from test_one_command_v13 import _write_fixture
+
+
+def _write_complete_fixture(root: Path, project_id: str) -> None:
+    _write_fixture(root, complete_request=True)
+    project_path = root / "project.yaml"
+    payload = yaml.safe_load(project_path.read_text(encoding="utf-8"))
+    payload["project_id"] = project_id
+    payload["name"] = f"{project_id} project"
+    payload["workcell_id"] = f"{project_id}-cell"
+    payload["robot_base_pose"] = {
+        "x": 0,
+        "y": 0,
+        "z": 0,
+        "roll": 0,
+        "pitch": 0,
+        "yaw": 0,
+    }
+    project_path.write_text(yaml.safe_dump(payload), encoding="utf-8")
+
+
+class OneCommandRegressionV141Tests(unittest.TestCase):
+    def test_three_consecutive_runs_use_immediately_previous_campaign(self) -> None:
+        with tempfile.TemporaryDirectory() as tmp:
+            root = Path(tmp) / "cell"
+            _write_complete_fixture(root, "repeat-cell")
+            db = Path(tmp) / "projects.db"
+            verifier = OneCommandVerifier(db)
+
+            results = [
+                verifier.verify(root, output_dir=Path(tmp) / f"run-{index}")
+                for index in range(1, 4)
+            ]
+            self.assertTrue(all(result.fat_artifact_id for result in results))
+            self.assertTrue(all(result.campaign_artifact_id for result in results))
+
+            repository = CommercialProjectRepository(db)
+            campaigns = repository.list_artifacts(
+                "repeat-cell",
+                artifact_type="commercial_campaign",
+            )
+            regressions = repository.list_artifacts(
+                "repeat-cell",
+                artifact_type="commercial_regression",
+            )
+            fat_reports = repository.list_artifacts(
+                "repeat-cell",
+                artifact_type="fat_report",
+            )
+
+            self.assertEqual(len(campaigns), 3)
+            self.assertEqual(len(regressions), 2)
+            self.assertEqual(len(fat_reports), 3)
+            self.assertEqual(
+                regressions[0].payload["baseline_campaign"],
+                campaigns[0].artifact_id,
+            )
+            self.assertEqual(
+                regressions[0].payload["current_campaign"],
+                campaigns[1].artifact_id,
+            )
+            self.assertEqual(
+                regressions[1].payload["baseline_campaign"],
+                campaigns[1].artifact_id,
+            )
+            self.assertEqual(
+                regressions[1].payload["current_campaign"],
+                campaigns[2].artifact_id,
+            )
+
+    def test_campaign_history_is_scoped_to_each_project(self) -> None:
+        with tempfile.TemporaryDirectory() as tmp:
+            db = Path(tmp) / "projects.db"
+            first = Path(tmp) / "first"
+            second = Path(tmp) / "second"
+            _write_complete_fixture(first, "first-cell")
+            _write_complete_fixture(second, "second-cell")
+            verifier = OneCommandVerifier(db)
+
+            verifier.verify(first, output_dir=Path(tmp) / "first-run-1")
+            verifier.verify(first, output_dir=Path(tmp) / "first-run-2")
+            second_result = verifier.verify(second, output_dir=Path(tmp) / "second-run-1")
+            self.assertIsNotNone(second_result.fat_artifact_id)
+
+            repository = CommercialProjectRepository(db)
+            self.assertEqual(
+                len(repository.list_artifacts("first-cell", artifact_type="commercial_regression")),
+                1,
+            )
+            self.assertEqual(
+                len(repository.list_artifacts("second-cell", artifact_type="commercial_regression")),
+                0,
+            )
+            self.assertEqual(
+                len(repository.list_artifacts("second-cell", artifact_type="commercial_campaign")),
+                1,
+            )
+
+    def test_customer_cli_bounds_commercial_error_without_traceback(self) -> None:
+        stdout = io.StringIO()
+        stderr = io.StringIO()
+        with (
+            patch(
+                "devagent_physical_engine.user_cli.main",
+                side_effect=CommercialError("fat_report_regression_scope_mismatch"),
+            ),
+            redirect_stdout(stdout),
+            redirect_stderr(stderr),
+        ):
+            rc = runtime_main(["verify", "/tmp/customer", "--json"])
+
+        self.assertEqual(rc, 10)
+        self.assertEqual(stderr.getvalue(), "")
+        payload = json.loads(stdout.getvalue())
+        self.assertFalse(payload["passed"])
+        self.assertEqual(
+            payload["failure_code"],
+            "fat_report_regression_scope_mismatch",
+        )
+        self.assertFalse(payload["real_execution_allowed"])
+
+
+if __name__ == "__main__":
+    unittest.main()