feat(engine): 添加运行时上下文支持并重构工具迭代限制

添加 RuntimeContext 类用于捕获模型运行时的日期时间信息，
包括UTC时间、本地时间和时区信息，并在系统提示中显示这些信息。

同时增加最大上下文消息数和工具迭代次数的配置选项，
将验证服务从引擎加载器中移除，并更新相关的数据结构和接口。

BREAKING CHANGE: 移除了验证服务，相关字段被替换为证据状态和接受状态。

- 添加 RuntimeContext 类和相关渲染方法
- 增加 max_context_messages 和 max_tool_iterations 配置
- 移除 ValidationService 相关代码
- 更新消息记录中的验证状态字段
- 添加原始工具调用检测和回退处理

app-instance/backend/beaver/tasks/__init__.py

@@ -6,7 +6,6 @@ from .planner import TaskExecutionPlan, TaskExecutionPlanner
 from .router import MainAgentRouter
 from .service import TaskService
 from .skill_resolver import SkillResolutionReport, TaskSkillResolver
-from .validation import ValidationService
 
 __all__ = [
     "EvidenceBuilder",
@@ -24,6 +23,5 @@ __all__ = [
     "ToolEvidence",
     "ValidationResult",
     "ValidationStatus",
-    "ValidationService",
     "render_task_evidence",
 ]

app-instance/backend/beaver/tasks/models.py

@@ -1,4 +1,4 @@
-"""Models for internal task tracking and validation."""
+"""Models for internal task tracking and user acceptance."""
 
 from __future__ import annotations
 
@@ -9,7 +9,12 @@ from typing import Any, Literal
 ValidationStatus = Literal["accepted", "rejected", "insufficient_evidence", "validator_error"]
 
 VALIDATION_STATUSES = {"accepted", "rejected", "insufficient_evidence", "validator_error"}
-TASK_OPEN_STATUSES = {"open", "running", "validating", "awaiting_feedback", "needs_review", "needs_revision"}
+TASK_OPEN_STATUSES = {"open", "running", "awaiting_acceptance", "needs_revision"}
+LEGACY_STATUS_MAP = {
+    "validating": "running",
+    "awaiting_feedback": "awaiting_acceptance",
+    "needs_review": "awaiting_acceptance",
+}
 
 
 @dataclass(slots=True)
@@ -113,11 +118,11 @@ class TaskRecord:
 
     @property
     def is_execution_active(self) -> bool:
-        return self.status in {"running", "validating"}
+        return self.status == "running"
 
     @property
     def requires_user_action(self) -> bool:
-        return self.status in {"awaiting_feedback", "needs_review", "needs_revision"}
+        return self.status in {"awaiting_acceptance", "needs_revision"}
 
     def to_dict(self) -> dict[str, Any]:
         return {
@@ -137,6 +142,7 @@ class TaskRecord:
             "satisfaction": self.satisfaction,
             "run_ids": list(self.run_ids),
             "skill_names": list(self.skill_names),
+            "acceptance": list(self.feedback),
             "feedback": list(self.feedback),
             "validation_result": self.validation_result,
             "metadata": dict(self.metadata),
@@ -152,7 +158,7 @@ class TaskRecord:
             goal=str(payload.get("goal") or payload.get("description") or ""),
             constraints=[str(item) for item in payload.get("constraints") or []],
             priority=int(payload.get("priority", 0) or 0),
-            status=str(payload.get("status") or "open"),
+            status=LEGACY_STATUS_MAP.get(str(payload.get("status") or "open"), str(payload.get("status") or "open")),
             creator=str(payload.get("creator") or "main-agent"),
             created_at=str(payload.get("created_at") or ""),
             updated_at=str(payload.get("updated_at") or ""),
@@ -161,7 +167,11 @@ class TaskRecord:
             satisfaction=_optional_float(payload.get("satisfaction")),
             run_ids=[str(item) for item in payload.get("run_ids") or []],
             skill_names=[str(item) for item in payload.get("skill_names") or []],
-            feedback=[dict(item) for item in payload.get("feedback") or [] if isinstance(item, dict)],
+            feedback=[
+                _normalize_acceptance_entry(dict(item))
+                for item in (payload.get("acceptance") or payload.get("feedback") or [])
+                if isinstance(item, dict)
+            ],
             validation_result=dict(payload["validation_result"]) if isinstance(payload.get("validation_result"), dict) else None,
             metadata=dict(payload.get("metadata") or {}),
         )
@@ -226,3 +236,13 @@ def _optional_float(value: Any) -> float | None:
     if value in (None, ""):
         return None
     return float(value)
+
+
+def _normalize_acceptance_entry(entry: dict[str, Any]) -> dict[str, Any]:
+    if entry.get("acceptance_type") is None and entry.get("feedback_type") is not None:
+        feedback_type = str(entry.get("feedback_type") or "")
+        entry["acceptance_type"] = "accept" if feedback_type == "satisfied" else feedback_type
+    if entry.get("feedback_type") is None and entry.get("acceptance_type") is not None:
+        acceptance_type = str(entry.get("acceptance_type") or "")
+        entry["feedback_type"] = "satisfied" if acceptance_type == "accept" else acceptance_type
+    return entry

app-instance/backend/beaver/tasks/planner.py

@@ -10,7 +10,7 @@ from typing import Any, Literal
 from beaver.coordinator.models import AgentDescriptor, ExecutionGraph, ExecutionNode
 from beaver.engine.providers import ProviderBundle
 
-from .models import TaskRecord, ValidationResult
+from .models import TaskRecord
 from .skill_resolver import SkillResolutionReport, TaskSkillResolver
 
 
@@ -76,7 +76,6 @@ class TaskExecutionPlanner:
         task: TaskRecord,
         user_message: str,
         attempt_index: int,
-        latest_validation: ValidationResult | None = None,
         provider_bundle: ProviderBundle | None = None,
         timeout_seconds: float = 30.0,
     ) -> TaskExecutionPlan:
@@ -105,7 +104,6 @@ class TaskExecutionPlanner:
                                 task=task,
                                 user_message=user_message,
                                 attempt_index=attempt_index,
-                                latest_validation=latest_validation,
                             ),
                         },
                     ],
@@ -230,14 +228,10 @@ class TaskExecutionPlanner:
         task: TaskRecord,
         user_message: str,
         attempt_index: int,
-        latest_validation: ValidationResult | None,
     ) -> str:
-        validation_note = ""
-        if latest_validation is not None:
-            validation_note = (
-                "\nPrevious validation issues:\n"
-                + json.dumps(latest_validation.to_dict(), ensure_ascii=False)
-            )
+        history_note = ""
+        if task.feedback:
+            history_note = "\nRelevant task history:\n" + json.dumps(task.feedback[-5:], ensure_ascii=False)
         return (
             "Decide execution mode for this internal Task attempt.\n"
             "Use mode=team only when independent research, review, implementation slices, or staged checks "
@@ -254,7 +248,7 @@ class TaskExecutionPlanner:
             f"Task goal:\n{task.goal}\n\n"
             f"Current user request:\n{user_message}\n\n"
             f"Attempt index: {attempt_index}\n"
-            f"{validation_note}"
+            f"{history_note}"
         )
 
     @staticmethod

app-instance/backend/beaver/tasks/service.py

@@ -7,7 +7,7 @@ from pathlib import Path
 from typing import Any
 from uuid import uuid4
 
-from .models import TaskEvent, TaskRecord, ValidationResult
+from .models import TaskEvent, TaskRecord
 from .store import TaskStore
 
 
@@ -105,38 +105,70 @@ class TaskService:
         for name in skill_names or []:
             if name not in task.skill_names:
                 task.skill_names.append(name)
+        task.status = "awaiting_acceptance"
         task.updated_at = self._now()
         self.store.upsert_task(task)
         self._event(task, "run_completed", run_id=run_id, payload={"skill_names": skill_names or []})
+        self._event(task, "evidence_recorded", run_id=run_id, payload={"skill_names": skill_names or []})
         return task
 
-    def record_validation(
+    def add_acceptance(
         self,
         task_id: str,
-        run_id: str,
-        validation: ValidationResult,
         *,
-        final_attempt: bool = True,
-        has_usable_answer: bool = True,
+        acceptance_type: str,
+        comment: str | None = None,
+        run_id: str | None = None,
     ) -> TaskRecord:
         task = self._require(task_id)
         now = self._now()
-        if validation.status == "accepted":
-            task.status = "awaiting_feedback"
-        elif validation.status in {"insufficient_evidence", "validator_error"}:
-            task.status = "needs_review"
-        elif validation.status == "rejected" and not final_attempt:
+        normalized = normalize_acceptance_type(acceptance_type)
+        matching_acceptance = any(
+            item.get("run_id") == run_id and item.get("acceptance_type") == normalized
+            for item in task.feedback
+        )
+        conflicting_acceptance = next(
+            (
+                item
+                for item in task.feedback
+                if item.get("run_id") == run_id and item.get("acceptance_type") != normalized
+            ),
+            None,
+        )
+        if conflicting_acceptance is not None:
+            raise ValueError(
+                f"Acceptance for run_id={run_id!r} was already recorded as "
+                f"{conflicting_acceptance.get('acceptance_type')!r}"
+            )
+        if task.status in {"closed", "abandoned"} and not matching_acceptance:
+            raise ValueError(f"Task {task.task_id} is already finalized as {task.status!r}")
+        if matching_acceptance:
+            return task
+
+        entry = {
+            "acceptance_type": normalized,
+            "feedback_type": "satisfied" if normalized == "accept" else normalized,
+            "comment": comment or "",
+            "run_id": run_id,
+            "created_at": now,
+        }
+        task.feedback.append(entry)
+        if normalized == "revise":
             task.status = "needs_revision"
-        elif validation.status == "rejected" and has_usable_answer:
-            task.status = "needs_review"
-        else:
-            task.status = "failed"
+        elif normalized == "abandon":
+            task.status = "abandoned"
             task.closed_at = now
-            task.close_reason = "automatic validation rejected the final attempt"
+            task.close_reason = comment or "abandoned"
+        elif normalized == "accept":
+            task.status = "closed"
+            task.closed_at = now
+            task.close_reason = "accepted"
+            task.satisfaction = 1.0
+            if run_id:
+                task.metadata["final_accepted_run_id"] = run_id
         task.updated_at = now
-        task.validation_result = validation.to_dict()
         self.store.upsert_task(task)
-        self._event(task, "validated", run_id=run_id, payload=validation.to_dict())
+        self._event(task, f"acceptance_{normalized}", run_id=run_id, payload=entry)
         return task
 
     def add_feedback(
@@ -147,52 +179,12 @@ class TaskService:
         comment: str | None = None,
         run_id: str | None = None,
     ) -> TaskRecord:
-        task = self._require(task_id)
-        now = self._now()
-        matching_feedback = any(
-            item.get("run_id") == run_id and item.get("feedback_type") == feedback_type
-            for item in task.feedback
+        return self.add_acceptance(
+            task_id,
+            acceptance_type=feedback_type,
+            comment=comment,
+            run_id=run_id,
         )
-        conflicting_feedback = next(
-            (
-                item
-                for item in task.feedback
-                if item.get("run_id") == run_id and item.get("feedback_type") != feedback_type
-            ),
-            None,
-        )
-        if conflicting_feedback is not None:
-            raise ValueError(
-                f"Feedback for run_id={run_id!r} was already recorded as "
-                f"{conflicting_feedback.get('feedback_type')!r}"
-            )
-        if task.status in {"closed", "abandoned"} and not matching_feedback:
-            raise ValueError(f"Task {task.task_id} is already finalized as {task.status!r}")
-        if matching_feedback:
-            return task
-
-        entry = {
-            "feedback_type": feedback_type,
-            "comment": comment or "",
-            "run_id": run_id,
-            "created_at": now,
-        }
-        task.feedback.append(entry)
-        if feedback_type == "revise":
-            task.status = "needs_revision"
-        elif feedback_type == "abandon":
-            task.status = "abandoned"
-            task.closed_at = now
-            task.close_reason = comment or "abandoned"
-        elif feedback_type == "satisfied":
-            task.status = "closed"
-            task.closed_at = now
-            task.close_reason = "satisfied"
-            task.satisfaction = 1.0
-        task.updated_at = now
-        self.store.upsert_task(task)
-        self._event(task, f"feedback_{feedback_type}", run_id=run_id, payload=entry)
-        return task
 
     def close_task(self, task_id: str, *, reason: str = "closed") -> TaskRecord:
         task = self._require(task_id)
@@ -267,3 +259,12 @@ def short_task_title(text: str) -> str:
     if len(words) <= 4:
         return cleaned[:40]
     return " ".join(words[:4])[:40]
+
+
+def normalize_acceptance_type(value: str) -> str:
+    normalized = (value or "").strip().lower()
+    if normalized == "satisfied":
+        return "accept"
+    if normalized not in {"accept", "revise", "abandon"}:
+        raise ValueError("acceptance_type must be one of: accept, revise, abandon")
+    return normalized

app-instance/backend/beaver/tasks/validation.py

@@ -1,154 +0,0 @@
-"""Automatic validation for internal Task mode."""
-
-from __future__ import annotations
-
-import json
-from typing import Any
-
-from beaver.engine.providers import ProviderBundle
-
-from .models import TaskRecord, ValidationResult
-
-
-class ValidationService:
-    async def validate_task_result(
-        self,
-        *,
-        task: TaskRecord,
-        user_message: str,
-        final_output: str,
-        evidence_packet: Any | None = None,
-        evidence_text: str = "",
-        transcript_excerpt: str = "",
-        tool_summaries: list[str] | None = None,
-        team_summaries: list[str] | None = None,
-        provider_bundle: ProviderBundle | None = None,
-    ) -> ValidationResult:
-        provider = None
-        model = None
-        if provider_bundle is not None:
-            provider = provider_bundle.auxiliary_provider or provider_bundle.main_provider
-            runtime = provider_bundle.auxiliary_runtime or provider_bundle.main_runtime
-            model = getattr(runtime, "model", None)
-        if provider is not None:
-            try:
-                return await self._validate_with_provider(
-                    provider=provider,
-                    model=model,
-                    task=task,
-                    user_message=user_message,
-                    final_output=final_output,
-                    evidence_text=evidence_text,
-                    transcript_excerpt=transcript_excerpt,
-                    tool_summaries=tool_summaries or [],
-                    team_summaries=team_summaries or [],
-                )
-            except Exception as exc:
-                return ValidationResult(
-                    status="validator_error",
-                    score=0.0,
-                    issues=[f"Validator failed: {exc}"],
-                    evidence_gaps=["Automatic validation failed before producing a reliable decision."],
-                    missing_requirements=["User review is required because automatic validation failed."],
-                    recommended_revision_prompt=(
-                        "Review the answer and evidence, then decide whether to revise or accept it."
-                    ),
-                    validator="llm_error",
-                )
-        return self._heuristic_validate(final_output)
-
-    async def _validate_with_provider(
-        self,
-        *,
-        provider: Any,
-        model: str | None,
-        task: TaskRecord,
-        user_message: str,
-        final_output: str,
-        evidence_text: str,
-        transcript_excerpt: str,
-        tool_summaries: list[str],
-        team_summaries: list[str],
-    ) -> ValidationResult:
-        legacy_context = "" if evidence_text else (
-            f"Transcript excerpt:\n{transcript_excerpt}\n\n"
-            f"Tool summaries:\n{json.dumps(tool_summaries, ensure_ascii=False)}\n\n"
-            f"Team summaries:\n{json.dumps(team_summaries, ensure_ascii=False)}\n\n"
-        )
-        prompt = (
-            "Validate whether the assistant output satisfies the task. "
-            "Return only compact JSON with keys: passed, score, issues, "
-            "missing_requirements, recommended_revision_prompt.\n\n"
-            f"Task goal:\n{task.goal}\n\n"
-            f"Current user request:\n{user_message}\n\n"
-            f"Evidence packet:\n{evidence_text}\n\n"
-            f"{legacy_context}"
-            f"Assistant final output:\n{final_output}"
-        )
-        response = await provider.chat(
-            messages=[
-                {"role": "system", "content": "You are a strict task result validator."},
-                {"role": "user", "content": prompt},
-            ],
-            tools=None,
-            model=model,
-            max_tokens=4096,
-            temperature=0.0,
-        )
-        payload = self._parse_json_object(response.content or "")
-        status = payload.get("status")
-        if status not in {"accepted", "rejected", "insufficient_evidence", "validator_error"}:
-            status = (
-                "accepted"
-                if payload.get("passed") and float(payload.get("score", 0.0) or 0.0) >= 0.75
-                else "rejected"
-            )
-        return ValidationResult(
-            status=status,
-            score=max(0.0, min(1.0, float(payload.get("score", 0.0) or 0.0))),
-            issues=[str(item) for item in payload.get("issues") or []],
-            missing_requirements=[str(item) for item in payload.get("missing_requirements") or []],
-            evidence_gaps=[str(item) for item in payload.get("evidence_gaps") or []],
-            recommended_revision_prompt=str(payload.get("recommended_revision_prompt") or ""),
-            validator="llm",
-        )
-
-    @staticmethod
-    def _heuristic_validate(final_output: str) -> ValidationResult:
-        text = final_output.strip()
-        if not text:
-            return ValidationResult(
-                passed=False,
-                score=0.0,
-                issues=["Assistant output is empty."],
-                missing_requirements=["A non-empty result is required."],
-                recommended_revision_prompt="Produce a complete, non-empty answer for the task.",
-                validator="heuristic",
-            )
-        lowered = text.lower()
-        if "run failed before completion" in lowered or "tool loop stopped" in lowered:
-            return ValidationResult(
-                passed=False,
-                score=0.35,
-                issues=["The run did not complete cleanly."],
-                missing_requirements=["A successful final result is required."],
-                recommended_revision_prompt="Retry the task and address the failure before returning the final answer.",
-                validator="heuristic",
-            )
-        return ValidationResult(passed=True, score=0.85, validator="heuristic")
-
-    @staticmethod
-    def _parse_json_object(text: str) -> dict[str, Any]:
-        cleaned = text.strip()
-        if cleaned.startswith("```"):
-            cleaned = cleaned.strip("`")
-            if cleaned.lower().startswith("json"):
-                cleaned = cleaned[4:].strip()
-        start = cleaned.find("{")
-        end = cleaned.rfind("}")
-        if start >= 0 and end >= start:
-            cleaned = cleaned[start : end + 1]
-        payload = json.loads(cleaned)
-        if not isinstance(payload, dict):
-            raise ValueError("validator response must be a JSON object")
-        return payload