forgexa-cli 1.57.5__tar.gz → 1.57.6__tar.gz

{forgexa_cli-1.57.5 → forgexa_cli-1.57.6}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: forgexa-cli
-Version: 1.57.5
+Version: 1.57.6
 Summary: Forgexa CLI — command-line client and AI agent runtime for the Forgexa platform
 Author-email: Jason Sun <dev.winds@gmail.com>
 License-Expression: MIT

{forgexa_cli-1.57.5 → forgexa_cli-1.57.6}/forgexa_cli/__init__.py

@@ -1,2 +1,2 @@
 """forgexa-cli — Forgexa command-line client."""
-__version__ = "1.57.5"
+__version__ = "1.57.6"

{forgexa_cli-1.57.5 → forgexa_cli-1.57.6}/forgexa_cli/daemon.py

@@ -1150,7 +1150,7 @@ except (ImportError, ModuleNotFoundError):
 # DAEMON_VERSION is the protocol/logic version of the daemon code.
 # Kept in sync with pyproject.toml version via bump-version.sh.
 # CLIENT_TYPE identifies which packaging/distribution this daemon runs in.
-DAEMON_VERSION = "1.57.5"
+DAEMON_VERSION = "1.57.6"
 
 
 def _detect_client_type() -> str:
@@ -1736,6 +1736,7 @@ def _filter_local_bind_task_result(result: "TaskResult") -> None:
     result.stdout = ""
     result.stderr = ""
     result.observations = []
+    result.metrics.pop("validation_issues", None)
     critic_review = result.critic_review
     if isinstance(critic_review, dict) and isinstance(critic_review.get("result"), dict):
         result.critic_review = {
@@ -1755,8 +1756,11 @@ def _filter_local_bind_task_result(result: "TaskResult") -> None:
         }
     if result.error:
         result.error = (
-            "Local task failed. Review this task in Forgexa Desktop on the "
-            "running device."
+            "Task verification is incomplete: acceptance criteria remain uncovered. "
+            "Review the validation log in Forgexa Desktop; resolve the criteria and "
+            "commit or stash local changes before retrying."
+            if result.failure_code == "acceptance_criteria_uncovered"
+            else "Local task failed. Review this task in Forgexa Desktop on the running device."
         )
 
 def _prepare_ai_job_output(
@@ -5726,7 +5730,9 @@ class ProcessManager:
     @staticmethod
     def agent_failure_kind(result: "TaskResult") -> str | None:
         """Classify a fallback-eligible failure without conflating its cause."""
-        if result.failure_code in {"all_agents_rate_limited", "all_agents_unavailable"}:
+        if result.failure_code in {
+            "all_agents_rate_limited", "all_agents_unavailable", "acceptance_criteria_uncovered",
+        }:
             return None
         if result.failure_code == "agent_platform_unsupported":
             return "agent_unavailable"
@@ -13052,7 +13058,7 @@ class RuntimeDaemon:
                 #   are *optional*: if they exist they are validated, but their
                 #   absence is not an error (the agent only writes regression
                 #   tests + test-report.md).
-                _skip_test_artifacts = False
+                _skip_test_artifacts = req_type == "task"
                 _requires_structured_artifacts = False
                 # Scope-adaptive pipeline (Phase 1): a content_only change
                 # scope (analysis-predicted, machine-verified server-side)
@@ -13074,19 +13080,28 @@ class RuntimeDaemon:
                     # importable (standalone vendored daemon execution).
                     _skip_test_artifacts = True
 
-                # The type-profile test_coverage skip (e.g. "task") is a
-                # lightweight-verify default — NOT an exemption from the e2e
-                # grounding chain or from a frozen enforce-mode artifact
-                # contract that was injected into the prompt. Both apply to
-                # every requirement type: SI-1616 (task verify) failed the e2e
-                # evidence chain because the skip semantics diverged between
-                # server daemons (profile skip applied) and standalone desktop
-                # daemons (profile import fails → nothing skipped).
+                # Keep e2e evidence validation active for any requirement type.
+                # An enforced policy requires artifacts only when the analysis
+                # contract expects test intent or the analysis supplied one.
+                _analysis_dir = (
+                    input_data.get("analysis_output_dir")
+                    or (input_data.get("context") or {}).get("analysis_output_dir")
+                    or ""
+                )
+                _has_test_intent = bool(_analysis_dir) and (
+                    workspace_path / str(_analysis_dir) / "test-intent.json"
+                ).is_file()
                 if (
                     _skip_test_artifacts
                     and _change_scope != "content_only"
                     and (
-                        (qa_policy or {}).get("enforcement") == "enforce"
+                        (
+                            (qa_policy or {}).get("enforcement") == "enforce"
+                            and (
+                                "test-intent.json" in _get_analysis_outputs_for_type(req_type, _change_scope)
+                                or _has_test_intent
+                            )
+                        )
                         or _testing_intents_have_e2e(workspace_path, input_data)
                     )
                 ):
@@ -13261,15 +13276,30 @@ class RuntimeDaemon:
                         ))
 
                     # --- coverage-matrix.json validation ---
+                    if req_type == "task":
+                        result.metrics.pop("acceptance_criteria_manual_review_required", None)
                     cm_path = base / "coverage-matrix.json"
                     if cm_path.exists():
                         try:
-                            cm_data = _json.loads(cm_path.read_text(encoding="utf-8"))
-                            ac_list = cm_data.get("acceptance_criteria", [])
-                            uncovered = [ac for ac in ac_list if ac.get("status") != "covered"]
-                            if uncovered:
-                                ids = ", ".join(ac.get("id", "?") for ac in uncovered[:5])
-                                issues.append(f"Uncovered acceptance criteria in coverage-matrix.json: {ids}")
+                            cm_content = cm_path.read_text(encoding="utf-8")
+                            cm_data = _json.loads(cm_content)
+                            if req_type == "task":
+                                coverage = runtime_evidence.parse_acceptance_criteria_coverage_artifacts(
+                                    _json.dumps(_tc_data) if _tc_data is not None else None,
+                                    cm_content,
+                                )
+                                links = coverage.get("coverage_links") or []
+                                if coverage.get("collection_status") != "completed" or not links:
+                                    reason = coverage.get("failure_reason") or "coverage_matrix_empty"
+                                    issues.append(f"coverage-matrix.json has invalid acceptance criteria: {reason}")
+                                else:
+                                    result.metrics["acceptance_criteria_manual_review_required"] = True
+                            else:
+                                ac_list = cm_data.get("acceptance_criteria", [])
+                                uncovered = [ac for ac in ac_list if ac.get("status") != "covered"]
+                                if uncovered:
+                                    ids = ", ".join(ac.get("id", "?") for ac in uncovered[:5])
+                                    issues.append(f"Uncovered acceptance criteria in coverage-matrix.json: {ids}")
                         except (_json.JSONDecodeError, UnicodeDecodeError) as e:
                             issues.append(f"coverage-matrix.json is not valid JSON: {e}")
                     elif _requires_structured_artifacts or _requires_e2e_artifacts:
@@ -14693,6 +14723,29 @@ class RuntimeDaemon:
                 )
                 return result  # All validations passed
 
+            if getattr(task, "workspace_resolution", None) == "local_bind":
+                _append_local_bind_output_log(task.task_id, [
+                    "[daemon] Output validation found these issues:",
+                    *[f"[daemon] {issue}" for issue in issues],
+                ])
+
+            if (
+                task.node_type == "testing"
+                and (task.input_data or {}).get("requirement_type") == "task"
+                and all(
+                    issue.startswith("Uncovered acceptance criteria in coverage-matrix.json:")
+                    for issue in issues
+                )
+            ):
+                result.status = "failed"
+                result.failure_code = "acceptance_criteria_uncovered"
+                result.error = f"Task verification is incomplete: {issues[0]}"
+                result.metrics["validation_issues"] = issues
+                self._record_testing_telemetry(
+                    workspace_path, task, result, retries_used=attempt
+                )
+                return result
+
             issues_text = "\n".join(f"- {iss}" for iss in issues)
             logger.warning(
                 "Validation gate: %d issues found for task %s (attempt %d/%d):\n%s",
@@ -14786,6 +14839,27 @@ class RuntimeDaemon:
         )
         remaining = self._validate_outputs(workspace_path, task, result)
         if remaining:
+            if getattr(task, "workspace_resolution", None) == "local_bind":
+                _append_local_bind_output_log(task.task_id, [
+                    "[daemon] Output validation remains incomplete:",
+                    *[f"[daemon] {issue}" for issue in remaining],
+                ])
+            if (
+                task.node_type == "testing"
+                and (task.input_data or {}).get("requirement_type") == "task"
+                and all(
+                    issue.startswith("Uncovered acceptance criteria in coverage-matrix.json:")
+                    for issue in remaining
+                )
+            ):
+                result.status = "failed"
+                result.failure_code = "acceptance_criteria_uncovered"
+                result.error = f"Task verification is incomplete: {remaining[0]}"
+                result.metrics["validation_issues"] = remaining
+                self._record_testing_telemetry(
+                    workspace_path, task, result, retries_used=max_retries
+                )
+                return result
             # Distinguish blocking artifact-contract failures from advisory syntax
             # warnings. Evidence coverage and references are release-blocking even
             # when the file itself exists.
@@ -14799,6 +14873,7 @@ class RuntimeDaemon:
                 "not found",
                 "is empty",
                 "not valid json",
+                "invalid acceptance criteria",
                 "case_evidence",
                 "test-evidence.json has no",
                 "evidence item",

{forgexa_cli-1.57.5 → forgexa_cli-1.57.6}/forgexa_cli.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: forgexa-cli
-Version: 1.57.5
+Version: 1.57.6
 Summary: Forgexa CLI — command-line client and AI agent runtime for the Forgexa platform
 Author-email: Jason Sun <dev.winds@gmail.com>
 License-Expression: MIT

{forgexa_cli-1.57.5 → forgexa_cli-1.57.6}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "forgexa-cli"
-version = "1.57.5"
+version = "1.57.6"
 description = "Forgexa CLI — command-line client and AI agent runtime for the Forgexa platform"
 requires-python = ">=3.9"
 license = "MIT"