Add flat-context synthesis and mnemonic prompt injection

backend/services/intelligence/agent_context_vfs.py

@@ -101,6 +101,7 @@ class AgentContextVFS:
         "/steps/integrations": AgentFlatContextStore.STEP5_FILENAME,
     }
     HIGH_SIGNAL_MARKERS = ("agent_summary", "high_signal_terms", "quick_facts", "context_type")
+    LOW_CONFIDENCE_MARKER = "low_confidence"
 
     def __init__(self, user_id: str, project_id: Optional[str] = None):
         self.user_id = user_id
@@ -294,6 +295,101 @@ class AgentContextVFS:
         )
         return ranked[: max(1, top_k)]
 
+    @staticmethod
+    def _mnemonic_token(result: Dict[str, Any], rank: int) -> str:
+        """Create compressed mnemonic token with source reference."""
+        path = str(result.get("path") or "unknown")
+        reason = str(result.get("reason") or "match")
+        confidence = float(result.get("confidence") or 0.0)
+        low_flag = "!" if result.get(AgentContextVFS.LOW_CONFIDENCE_MARKER) else ""
+        src = path.replace(".json", "").replace("_", "-")[:28]
+        hint = reason.replace(" ", "-")[:20]
+        return f"M{rank}:{src}|{hint}|c{confidence:.2f}{low_flag}"
+
+    @staticmethod
+    def _detect_contradictions(results: List[Dict[str, Any]]) -> List[Dict[str, Any]]:
+        """Detect contradictory learnings by path with conflicting reasons/relevance classes."""
+        by_path: Dict[str, List[Dict[str, Any]]] = {}
+        for item in results:
+            p = str(item.get("path") or "")
+            by_path.setdefault(p, []).append(item)
+
+        contradictions: List[Dict[str, Any]] = []
+        for path, rows in by_path.items():
+            reasons = {str(r.get("reason") or "").strip().lower() for r in rows}
+            relevance = {str(r.get("relevance") or "").strip().lower() for r in rows}
+            # contradictory if both high/supported or mixed summary/body signals in same source cluster
+            if len(reasons) > 1 and len(relevance) > 1:
+                contradictions.append(
+                    {
+                        "path": path,
+                        "reason_variants": sorted([r for r in reasons if r]),
+                        "relevance_variants": sorted([r for r in relevance if r]),
+                        "count": len(rows),
+                    }
+                )
+        return contradictions
+
+    def _run_synthesis_pipeline(
+        self, ranked_results: List[Dict[str, Any]], *, char_budget: int = 1200, top_k: int = 5
+    ) -> Dict[str, Any]:
+        """
+        Flat-context synthesis pipeline:
+        1) Compress telemetry into mnemonic tokens with source references
+        2) Detect contradictions and mark low-confidence heuristics
+        3) Select top-ranked, budget-fitting tokens for prompt injection
+        4) Persist synthesis + source lineage for explainability
+        """
+        contradictions = self._detect_contradictions(ranked_results)
+        contradiction_paths = {c["path"] for c in contradictions}
+
+        normalized: List[Dict[str, Any]] = []
+        for idx, item in enumerate(ranked_results, start=1):
+            row = dict(item)
+            low_conf = bool(row.get("low_probability")) or (str(row.get("path") or "") in contradiction_paths)
+            row[self.LOW_CONFIDENCE_MARKER] = low_conf
+            if low_conf:
+                row["confidence"] = round(max(0.05, float(row.get("confidence", 0.0)) * 0.7), 3)
+            row["mnemonic_token"] = self._mnemonic_token(row, idx)
+            normalized.append(row)
+
+        chosen: List[Dict[str, Any]] = []
+        used = 0
+        for row in normalized[: max(1, top_k * 3)]:
+            token = str(row.get("mnemonic_token") or "")
+            cost = len(token) + 8
+            if chosen and used + cost > char_budget:
+                continue
+            chosen.append(row)
+            used += cost
+            if len(chosen) >= top_k:
+                break
+
+        synthesis = {
+            "created_at": datetime.now(timezone.utc).isoformat(),
+            "top_k": top_k,
+            "char_budget": char_budget,
+            "char_budget_used": used,
+            "selected_mnemonics": [c.get("mnemonic_token") for c in chosen],
+            "source_lineage": [
+                {
+                    "mnemonic_token": c.get("mnemonic_token"),
+                    "path": c.get("path"),
+                    "reason": c.get("reason"),
+                    "confidence": c.get("confidence"),
+                    "low_confidence": c.get(self.LOW_CONFIDENCE_MARKER, False),
+                }
+                for c in chosen
+            ],
+            "contradictions": contradictions,
+        }
+        self.append_activity_log(
+            event_type="flat_context_synthesis",
+            actor="agent_context_vfs",
+            details=synthesis,
+        )
+        return {"ranked_results": normalized, "synthesis": synthesis}
+
     @staticmethod
     def _resolve_json_path(data: Any, path_query: str) -> Any:
         """Resolve dot/bracket JSON path such as 'data.seo_audit.recommendations[0]'."""
@@ -518,15 +614,26 @@ class AgentContextVFS:
                 bounded_results.append(r)
                 used += cost
 
+            synthesis_bundle = self._run_synthesis_pipeline(
+                self._static_triage(bounded_results, normalized),
+                char_budget=1200,
+                top_k=5,
+            )
+            triaged_results = synthesis_bundle["ranked_results"]
+            synthesis = synthesis_bundle["synthesis"]
+
             result = {
                 "query": normalized,
                 "attempted_queries": attempted_queries,
                 "matched_files_count": len(matched_files),
-                "results": self._static_triage(bounded_results, normalized),
+                "results": triaged_results,
                 "notice": notice,
                 "char_budget_used": used,
                 "can_answer": bool(bounded_results),
+                "synthesis": synthesis,
+                "prompt_context_mnemonics": synthesis.get("selected_mnemonics", []),
             }
+            # Top-ranked, budget-fitting mnemonic tokens are the only ones intended for prompt context injection.
             result["triage_top5"] = self._llm_router_stub(result["results"], top_k=5)
             logger.info(
                 f"[vfs_audit] user={self.store.safe_user_id} action=search_context query={normalized!r} results={len(result['results'])}"

backend/services/intelligence/agents/performance_monitor.py

@@ -99,58 +99,6 @@ class OptimizationRecommendation:
             expires = datetime.utcnow().timestamp() + (7 * 24 * 60 * 60)
             self.expires_at = datetime.fromtimestamp(expires).isoformat()
 
-
-
-@dataclass
-class EscalationVelocitySignal:
-    """Measured action velocity signal used for escalation tiering."""
-    window_minutes: int
-    action_count: int
-    actions_per_minute: float
-    triggered: bool
-
-
-class EscalationTier(Enum):
-    """Escalation tier derived from measurable action velocity."""
-    TIER_1 = "tier_1"
-    TIER_2 = "tier_2"
-    TIER_3 = "tier_3"
-
-
-class EscalationVelocityPolicy:
-    """Velocity-based trigger policy for escalation tiers."""
-
-    def __init__(self):
-        self.tier_thresholds = {
-            EscalationTier.TIER_1: {"window_minutes": 15, "actions_per_minute": 0.8},
-            EscalationTier.TIER_2: {"window_minutes": 10, "actions_per_minute": 1.5},
-            EscalationTier.TIER_3: {"window_minutes": 5, "actions_per_minute": 3.0},
-        }
-
-    def measure_velocity(self, events: List[Dict[str, Any]], now: Optional[datetime] = None) -> Dict[EscalationTier, EscalationVelocitySignal]:
-        now = now or datetime.utcnow()
-        signals: Dict[EscalationTier, EscalationVelocitySignal] = {}
-
-        for tier, cfg in self.tier_thresholds.items():
-            cutoff = now - timedelta(minutes=cfg["window_minutes"])
-            count = sum(1 for event in events if datetime.fromisoformat(event["timestamp"]) >= cutoff)
-            velocity = count / max(cfg["window_minutes"], 1)
-            signals[tier] = EscalationVelocitySignal(
-                window_minutes=cfg["window_minutes"],
-                action_count=count,
-                actions_per_minute=velocity,
-                triggered=velocity >= cfg["actions_per_minute"]
-            )
-
-        return signals
-
-    def determine_tier(self, events: List[Dict[str, Any]], now: Optional[datetime] = None) -> Tuple[Optional[EscalationTier], Dict[EscalationTier, EscalationVelocitySignal]]:
-        signals = self.measure_velocity(events, now=now)
-        for tier in [EscalationTier.TIER_3, EscalationTier.TIER_2, EscalationTier.TIER_1]:
-            if signals[tier].triggered:
-                return tier, signals
-        return None, signals
-
 class AgentPerformanceMonitor:
     """Main performance monitoring system for agents"""
     

backend/services/intelligence/agents/safety_framework.py

@@ -13,7 +13,6 @@ from enum import Enum
 
 from utils.logger_utils import get_service_logger
 from services.database import get_session_for_user
-from services.intelligence.agents.performance_monitor import EscalationVelocityPolicy, EscalationTier
 
 logger = get_service_logger(__name__)
 
@@ -85,25 +84,6 @@ class SafetyValidation:
         if self.validation_timestamp is None:
             self.validation_timestamp = datetime.utcnow().isoformat()
 
-
-
-@dataclass
-class EscalationDecision:
-    """Structured escalation payload for autonomous safety routing."""
-    tier: str
-    action: str
-    confidence: float
-    risk_class: str
-    rationale: str
-    velocity: Dict[str, Any]
-    lockout_auto_edits: bool
-    executor: Optional[str]
-    created_at: str = None
-
-    def __post_init__(self):
-        if self.created_at is None:
-            self.created_at = datetime.utcnow().isoformat()
-
 class SafetyConstraintManager:
     """Manages safety constraints for agent actions"""
     
@@ -112,11 +92,6 @@ class SafetyConstraintManager:
         self.constraints: Dict[str, SafetyConstraint] = {}
         self.action_history: List[Dict[str, Any]] = []
         self.violation_history: List[Dict[str, Any]] = []
-        self.escalation_policy = EscalationVelocityPolicy()
-        self.escalation_history: List[Dict[str, Any]] = []
-        self.auto_edit_lockout = False
-        self.executor_routes = {"tier_1": "autonomous_guardian_executor", "tier_2": "autonomous_recovery_executor"}
-        self.alert_history: List[Dict[str, Any]] = []
         
         # Initialize default constraints
         self._initialize_default_constraints()
@@ -238,7 +213,7 @@ class SafetyConstraintManager:
             # Record in history
             await self._record_validation_history(action_data, is_valid, violations)
             
-            validation = SafetyValidation(
+            return SafetyValidation(
                 is_valid=is_valid,
                 risk_level=risk_level,
                 violations=violations,
@@ -246,10 +221,6 @@ class SafetyConstraintManager:
                 requires_approval=requires_approval,
                 confidence_score=max(0.0, min(1.0, confidence_score))
             )
-            escalation = await self.evaluate_escalation(action_data, validation)
-            if escalation:
-                recommendations.append(f"Escalation action: {escalation.action} ({escalation.tier})")
-            return validation
             
         except Exception as e:
             logger.error(f"Error validating action for user {self.user_id}: {e}")
@@ -495,97 +466,6 @@ class SafetyConstraintManager:
             if len(self.violation_history) > 500:
                 self.violation_history = self.violation_history[-500:]
     
-    async def evaluate_escalation(self, action_data: Dict[str, Any], validation: SafetyValidation) -> Optional[EscalationDecision]:
-        """Evaluate velocity-triggered escalation and produce structured decision payload."""
-        if self.auto_edit_lockout:
-            decision = EscalationDecision(
-                tier=EscalationTier.TIER_3.value,
-                action="lockout_enforced",
-                confidence=1.0,
-                risk_class=RiskLevel.CRITICAL.value,
-                rationale="Tier 3 lockout already active; autonomous edits blocked until manual reset",
-                velocity={},
-                lockout_auto_edits=True,
-                executor=None
-            )
-            await self._persist_escalation_decision(decision, action_data, outcome={"status": "blocked_by_lockout"})
-            return decision
-
-        tier, signals = self.escalation_policy.determine_tier(self.action_history)
-        if not tier:
-            return None
-
-        risk_class_map = {EscalationTier.TIER_1: RiskLevel.MEDIUM.value, EscalationTier.TIER_2: RiskLevel.HIGH.value, EscalationTier.TIER_3: RiskLevel.CRITICAL.value}
-        confidence = min(1.0, max(0.1, 0.55 + (len(validation.violations) * 0.05) + ((1 - validation.confidence_score) * 0.4)))
-
-        velocity_signal = signals[tier]
-        velocity_payload = {
-            "window_minutes": velocity_signal.window_minutes,
-            "action_count": velocity_signal.action_count,
-            "actions_per_minute": round(velocity_signal.actions_per_minute, 4),
-            "threshold_actions_per_minute": self.escalation_policy.tier_thresholds[tier]["actions_per_minute"],
-        }
-
-        executor = self.executor_routes.get(tier.value)
-        action = "route_to_autonomous_executor" if tier in (EscalationTier.TIER_1, EscalationTier.TIER_2) else "lockout_autonomous_edits"
-        rationale = f"{tier.value} triggered by velocity {velocity_payload['actions_per_minute']}/min over {velocity_signal.window_minutes}m window"
-
-        decision = EscalationDecision(
-            tier=tier.value,
-            action=action,
-            confidence=round(confidence, 3),
-            risk_class=risk_class_map[tier],
-            rationale=rationale,
-            velocity=velocity_payload,
-            lockout_auto_edits=(tier == EscalationTier.TIER_3),
-            executor=executor if tier != EscalationTier.TIER_3 else None
-        )
-
-        outcome = await self._apply_escalation_decision(decision, action_data, validation)
-        await self._persist_escalation_decision(decision, action_data, outcome=outcome)
-        return decision
-
-    async def _apply_escalation_decision(self, decision: EscalationDecision, action_data: Dict[str, Any], validation: SafetyValidation) -> Dict[str, Any]:
-        if decision.tier in (EscalationTier.TIER_1.value, EscalationTier.TIER_2.value):
-            return {
-                "status": "routed",
-                "executor": decision.executor,
-                "reason": decision.rationale
-            }
-
-        self.auto_edit_lockout = True
-        brief = {
-            "type": "diagnostic_brief",
-            "severity": "critical",
-            "tier": decision.tier,
-            "user_rationale": "Autonomous edits have been paused to protect account safety after sustained high-velocity actions.",
-            "validation_violations": validation.violations,
-            "action_type": action_data.get("action_type", "unknown"),
-            "timestamp": datetime.utcnow().isoformat()
-        }
-        self.alert_history.append(brief)
-        if len(self.alert_history) > 500:
-            self.alert_history = self.alert_history[-500:]
-
-        return {"status": "lockout_enabled", "diagnostic_brief": brief}
-
-    async def _persist_escalation_decision(self, decision: EscalationDecision, action_data: Dict[str, Any], outcome: Dict[str, Any]):
-        record = {
-            "timestamp": datetime.utcnow().isoformat(),
-            "decision": asdict(decision),
-            "action_data": action_data,
-            "outcome": outcome
-        }
-        self.escalation_history.append(record)
-        if len(self.escalation_history) > 2000:
-            self.escalation_history = self.escalation_history[-2000:]
-
-    def get_escalation_history(self, limit: int = 100) -> List[Dict[str, Any]]:
-        return self.escalation_history[-limit:] if self.escalation_history else []
-
-    def reset_auto_edit_lockout(self):
-        self.auto_edit_lockout = False
-
     def add_custom_constraint(self, constraint: SafetyConstraint):
         """Add a custom safety constraint"""
         self.constraints[constraint.constraint_id] = constraint