@llm-dev-ops/agentics-cli 1.4.52 → 1.4.54

package/dist/bundled-agents/copilot-agents/functions/lib/services/agents/contracts/reflection-schemas.d.ts

@@ -0,0 +1,706 @@
+/**
+ * Reflection Agent Schemas
+ *
+ * Purpose: Analyze DecisionEvents to extract learning and quality signals
+ * Classification: POST_EXECUTION_ANALYSIS, QUALITY_ASSESSMENT
+ * decision_type: reflection_analysis
+ *
+ * Scope:
+ * - Evaluate outcomes
+ * - Identify gaps and inefficiencies
+ * - Produce improvement insights
+ *
+ * Must Never:
+ * - Modify behavior
+ * - Trigger retries
+ * - Apply optimizations
+ */
+import { z } from 'zod';
+/**
+ * Quality Signal - extracted quality indicators
+ */
+export declare const QualitySignalSchema: z.ZodObject<{
+    /** Unique identifier for this signal */
+    signal_id: z.ZodString;
+    /** Type of quality signal */
+    type: z.ZodEnum<["performance", "accuracy", "completeness", "consistency", "efficiency", "reliability"]>;
+    /** Signal value (0.0 - 1.0 normalized score) */
+    value: z.ZodNumber;
+    /** Human-readable description of the signal */
+    description: z.ZodString;
+    /** Evidence supporting this signal */
+    evidence: z.ZodDefault<z.ZodArray<z.ZodString, "many">>;
+    /** Severity level for negative signals */
+    severity: z.ZodOptional<z.ZodEnum<["info", "warning", "error", "critical"]>>;
+}, "strip", z.ZodTypeAny, {
+    value: number;
+    type: "performance" | "accuracy" | "completeness" | "consistency" | "efficiency" | "reliability";
+    description: string;
+    signal_id: string;
+    evidence: string[];
+    severity?: "error" | "critical" | "info" | "warning" | undefined;
+}, {
+    value: number;
+    type: "performance" | "accuracy" | "completeness" | "consistency" | "efficiency" | "reliability";
+    description: string;
+    signal_id: string;
+    severity?: "error" | "critical" | "info" | "warning" | undefined;
+    evidence?: string[] | undefined;
+}>;
+export type QualitySignal = z.infer<typeof QualitySignalSchema>;
+/**
+ * Learning Signal - extracted learning opportunities
+ */
+export declare const LearningSignalSchema: z.ZodObject<{
+    /** Unique identifier for this learning */
+    learning_id: z.ZodString;
+    /** Category of learning */
+    category: z.ZodEnum<["pattern", "anti_pattern", "optimization", "edge_case", "dependency", "constraint"]>;
+    /** Brief title for the learning */
+    title: z.ZodString;
+    /** Detailed description of the learning */
+    description: z.ZodString;
+    /** Confidence in this learning (0.0 - 1.0) */
+    confidence: z.ZodNumber;
+    /** Affected agents or components */
+    affected_agents: z.ZodDefault<z.ZodArray<z.ZodString, "many">>;
+    /** Recommendations for applying this learning (informational only) */
+    recommendations: z.ZodDefault<z.ZodArray<z.ZodString, "many">>;
+}, "strip", z.ZodTypeAny, {
+    confidence: number;
+    description: string;
+    learning_id: string;
+    category: "pattern" | "anti_pattern" | "optimization" | "edge_case" | "dependency" | "constraint";
+    title: string;
+    affected_agents: string[];
+    recommendations: string[];
+}, {
+    confidence: number;
+    description: string;
+    learning_id: string;
+    category: "pattern" | "anti_pattern" | "optimization" | "edge_case" | "dependency" | "constraint";
+    title: string;
+    affected_agents?: string[] | undefined;
+    recommendations?: string[] | undefined;
+}>;
+export type LearningSignal = z.infer<typeof LearningSignalSchema>;
+/**
+ * Gap Analysis - identified gaps or inefficiencies
+ */
+export declare const GapAnalysisSchema: z.ZodObject<{
+    /** Unique identifier for this gap */
+    gap_id: z.ZodString;
+    /** Type of gap */
+    type: z.ZodEnum<["coverage", "capability", "data", "process", "integration", "documentation"]>;
+    /** Brief title for the gap */
+    title: z.ZodString;
+    /** Detailed description of the gap */
+    description: z.ZodString;
+    /** Impact assessment */
+    impact: z.ZodEnum<["low", "medium", "high", "critical"]>;
+    /** Steps involved where gap was identified */
+    affected_steps: z.ZodDefault<z.ZodArray<z.ZodString, "many">>;
+    /** Evidence supporting gap identification */
+    evidence: z.ZodDefault<z.ZodArray<z.ZodString, "many">>;
+}, "strip", z.ZodTypeAny, {
+    type: "data" | "coverage" | "capability" | "process" | "integration" | "documentation";
+    description: string;
+    affected_steps: string[];
+    evidence: string[];
+    title: string;
+    gap_id: string;
+    impact: "low" | "medium" | "high" | "critical";
+}, {
+    type: "data" | "coverage" | "capability" | "process" | "integration" | "documentation";
+    description: string;
+    title: string;
+    gap_id: string;
+    impact: "low" | "medium" | "high" | "critical";
+    affected_steps?: string[] | undefined;
+    evidence?: string[] | undefined;
+}>;
+export type GapAnalysis = z.infer<typeof GapAnalysisSchema>;
+/**
+ * Outcome Evaluation - assessment of decision outcomes
+ */
+export declare const OutcomeEvaluationSchema: z.ZodObject<{
+    /** Original decision event being evaluated */
+    decision_ref: z.ZodString;
+    /** Agent that produced the decision */
+    agent_id: z.ZodString;
+    /** Decision type being evaluated */
+    decision_type: z.ZodString;
+    /** Overall outcome score (0.0 - 1.0) */
+    outcome_score: z.ZodNumber;
+    /** Assessment summary */
+    summary: z.ZodString;
+    /** Breakdown of outcome by dimension */
+    dimensions: z.ZodDefault<z.ZodArray<z.ZodObject<{
+        name: z.ZodString;
+        score: z.ZodNumber;
+        notes: z.ZodOptional<z.ZodString>;
+    }, "strip", z.ZodTypeAny, {
+        name: string;
+        score: number;
+        notes?: string | undefined;
+    }, {
+        name: string;
+        score: number;
+        notes?: string | undefined;
+    }>, "many">>;
+    /** Was the outcome as expected? */
+    met_expectations: z.ZodBoolean;
+    /** Deviation from expected outcome if any */
+    deviation_notes: z.ZodOptional<z.ZodString>;
+}, "strip", z.ZodTypeAny, {
+    agent_id: string;
+    decision_type: string;
+    decision_ref: string;
+    outcome_score: number;
+    summary: string;
+    dimensions: {
+        name: string;
+        score: number;
+        notes?: string | undefined;
+    }[];
+    met_expectations: boolean;
+    deviation_notes?: string | undefined;
+}, {
+    agent_id: string;
+    decision_type: string;
+    decision_ref: string;
+    outcome_score: number;
+    summary: string;
+    met_expectations: boolean;
+    dimensions?: {
+        name: string;
+        score: number;
+        notes?: string | undefined;
+    }[] | undefined;
+    deviation_notes?: string | undefined;
+}>;
+export type OutcomeEvaluation = z.infer<typeof OutcomeEvaluationSchema>;
+/**
+ * Reflection Agent Input Schema
+ */
+export declare const ReflectionInputSchema: z.ZodObject<{
+    /** DecisionEvents to analyze (1-100 events) */
+    decision_events: z.ZodArray<z.ZodObject<{
+        agent_id: z.ZodString;
+        agent_version: z.ZodString;
+        decision_type: z.ZodString;
+        inputs_hash: z.ZodString;
+        outputs: z.ZodUnknown;
+        confidence: z.ZodNumber;
+        constraints_applied: z.ZodArray<z.ZodString, "many">;
+        execution_ref: z.ZodString;
+        timestamp: z.ZodString;
+    }, "strip", z.ZodTypeAny, {
+        agent_id: string;
+        agent_version: string;
+        decision_type: string;
+        inputs_hash: string;
+        confidence: number;
+        constraints_applied: string[];
+        execution_ref: string;
+        timestamp: string;
+        outputs?: unknown;
+    }, {
+        agent_id: string;
+        agent_version: string;
+        decision_type: string;
+        inputs_hash: string;
+        confidence: number;
+        constraints_applied: string[];
+        execution_ref: string;
+        timestamp: string;
+        outputs?: unknown;
+    }>, "many">;
+    /** Optional context about the analysis scope */
+    context: z.ZodOptional<z.ZodObject<{
+        /** Focus areas for analysis */
+        focus_areas: z.ZodDefault<z.ZodArray<z.ZodEnum<["quality", "learning", "gaps", "outcomes", "all"]>, "many">>;
+        /** Time range context */
+        time_range: z.ZodOptional<z.ZodObject<{
+            from: z.ZodOptional<z.ZodString>;
+            to: z.ZodOptional<z.ZodString>;
+        }, "strip", z.ZodTypeAny, {
+            from?: string | undefined;
+            to?: string | undefined;
+        }, {
+            from?: string | undefined;
+            to?: string | undefined;
+        }>>;
+        /** Specific agents to focus on */
+        target_agents: z.ZodOptional<z.ZodArray<z.ZodString, "many">>;
+        /** Prior reflections to build upon */
+        prior_reflection_refs: z.ZodOptional<z.ZodArray<z.ZodString, "many">>;
+        /** Domain-specific context */
+        domain_context: z.ZodOptional<z.ZodRecord<z.ZodString, z.ZodString>>;
+    }, "strip", z.ZodTypeAny, {
+        focus_areas: ("quality" | "learning" | "gaps" | "outcomes" | "all")[];
+        time_range?: {
+            from?: string | undefined;
+            to?: string | undefined;
+        } | undefined;
+        target_agents?: string[] | undefined;
+        prior_reflection_refs?: string[] | undefined;
+        domain_context?: Record<string, string> | undefined;
+    }, {
+        focus_areas?: ("quality" | "learning" | "gaps" | "outcomes" | "all")[] | undefined;
+        time_range?: {
+            from?: string | undefined;
+            to?: string | undefined;
+        } | undefined;
+        target_agents?: string[] | undefined;
+        prior_reflection_refs?: string[] | undefined;
+        domain_context?: Record<string, string> | undefined;
+    }>>;
+    /** Analysis preferences */
+    preferences: z.ZodOptional<z.ZodObject<{
+        /** Minimum confidence threshold for signals (0.0 - 1.0) */
+        min_confidence: z.ZodDefault<z.ZodNumber>;
+        /** Maximum number of signals to return per category */
+        max_signals_per_category: z.ZodDefault<z.ZodNumber>;
+        /** Include detailed evidence */
+        include_evidence: z.ZodDefault<z.ZodBoolean>;
+        /** Generate cross-event correlations */
+        correlate_events: z.ZodDefault<z.ZodBoolean>;
+    }, "strip", z.ZodTypeAny, {
+        min_confidence: number;
+        max_signals_per_category: number;
+        include_evidence: boolean;
+        correlate_events: boolean;
+    }, {
+        min_confidence?: number | undefined;
+        max_signals_per_category?: number | undefined;
+        include_evidence?: boolean | undefined;
+        correlate_events?: boolean | undefined;
+    }>>;
+    /** Request ID for tracing */
+    request_id: z.ZodOptional<z.ZodString>;
+}, "strip", z.ZodTypeAny, {
+    decision_events: {
+        agent_id: string;
+        agent_version: string;
+        decision_type: string;
+        inputs_hash: string;
+        confidence: number;
+        constraints_applied: string[];
+        execution_ref: string;
+        timestamp: string;
+        outputs?: unknown;
+    }[];
+    preferences?: {
+        min_confidence: number;
+        max_signals_per_category: number;
+        include_evidence: boolean;
+        correlate_events: boolean;
+    } | undefined;
+    context?: {
+        focus_areas: ("quality" | "learning" | "gaps" | "outcomes" | "all")[];
+        time_range?: {
+            from?: string | undefined;
+            to?: string | undefined;
+        } | undefined;
+        target_agents?: string[] | undefined;
+        prior_reflection_refs?: string[] | undefined;
+        domain_context?: Record<string, string> | undefined;
+    } | undefined;
+    request_id?: string | undefined;
+}, {
+    decision_events: {
+        agent_id: string;
+        agent_version: string;
+        decision_type: string;
+        inputs_hash: string;
+        confidence: number;
+        constraints_applied: string[];
+        execution_ref: string;
+        timestamp: string;
+        outputs?: unknown;
+    }[];
+    preferences?: {
+        min_confidence?: number | undefined;
+        max_signals_per_category?: number | undefined;
+        include_evidence?: boolean | undefined;
+        correlate_events?: boolean | undefined;
+    } | undefined;
+    context?: {
+        focus_areas?: ("quality" | "learning" | "gaps" | "outcomes" | "all")[] | undefined;
+        time_range?: {
+            from?: string | undefined;
+            to?: string | undefined;
+        } | undefined;
+        target_agents?: string[] | undefined;
+        prior_reflection_refs?: string[] | undefined;
+        domain_context?: Record<string, string> | undefined;
+    } | undefined;
+    request_id?: string | undefined;
+}>;
+export type ReflectionInput = z.infer<typeof ReflectionInputSchema>;
+/**
+ * Reflection Agent Output Schema
+ */
+export declare const ReflectionOutputSchema: z.ZodObject<{
+    /** Unique reflection identifier */
+    reflection_id: z.ZodString;
+    /** Number of events analyzed */
+    events_analyzed: z.ZodNumber;
+    /** Agents covered in analysis */
+    agents_analyzed: z.ZodArray<z.ZodString, "many">;
+    /** Time range of analyzed events */
+    analysis_time_range: z.ZodObject<{
+        earliest: z.ZodString;
+        latest: z.ZodString;
+    }, "strip", z.ZodTypeAny, {
+        earliest: string;
+        latest: string;
+    }, {
+        earliest: string;
+        latest: string;
+    }>;
+    /** Outcome evaluations for each analyzed event */
+    outcome_evaluations: z.ZodArray<z.ZodObject<{
+        /** Original decision event being evaluated */
+        decision_ref: z.ZodString;
+        /** Agent that produced the decision */
+        agent_id: z.ZodString;
+        /** Decision type being evaluated */
+        decision_type: z.ZodString;
+        /** Overall outcome score (0.0 - 1.0) */
+        outcome_score: z.ZodNumber;
+        /** Assessment summary */
+        summary: z.ZodString;
+        /** Breakdown of outcome by dimension */
+        dimensions: z.ZodDefault<z.ZodArray<z.ZodObject<{
+            name: z.ZodString;
+            score: z.ZodNumber;
+            notes: z.ZodOptional<z.ZodString>;
+        }, "strip", z.ZodTypeAny, {
+            name: string;
+            score: number;
+            notes?: string | undefined;
+        }, {
+            name: string;
+            score: number;
+            notes?: string | undefined;
+        }>, "many">>;
+        /** Was the outcome as expected? */
+        met_expectations: z.ZodBoolean;
+        /** Deviation from expected outcome if any */
+        deviation_notes: z.ZodOptional<z.ZodString>;
+    }, "strip", z.ZodTypeAny, {
+        agent_id: string;
+        decision_type: string;
+        decision_ref: string;
+        outcome_score: number;
+        summary: string;
+        dimensions: {
+            name: string;
+            score: number;
+            notes?: string | undefined;
+        }[];
+        met_expectations: boolean;
+        deviation_notes?: string | undefined;
+    }, {
+        agent_id: string;
+        decision_type: string;
+        decision_ref: string;
+        outcome_score: number;
+        summary: string;
+        met_expectations: boolean;
+        dimensions?: {
+            name: string;
+            score: number;
+            notes?: string | undefined;
+        }[] | undefined;
+        deviation_notes?: string | undefined;
+    }>, "many">;
+    /** Extracted quality signals */
+    quality_signals: z.ZodArray<z.ZodObject<{
+        /** Unique identifier for this signal */
+        signal_id: z.ZodString;
+        /** Type of quality signal */
+        type: z.ZodEnum<["performance", "accuracy", "completeness", "consistency", "efficiency", "reliability"]>;
+        /** Signal value (0.0 - 1.0 normalized score) */
+        value: z.ZodNumber;
+        /** Human-readable description of the signal */
+        description: z.ZodString;
+        /** Evidence supporting this signal */
+        evidence: z.ZodDefault<z.ZodArray<z.ZodString, "many">>;
+        /** Severity level for negative signals */
+        severity: z.ZodOptional<z.ZodEnum<["info", "warning", "error", "critical"]>>;
+    }, "strip", z.ZodTypeAny, {
+        value: number;
+        type: "performance" | "accuracy" | "completeness" | "consistency" | "efficiency" | "reliability";
+        description: string;
+        signal_id: string;
+        evidence: string[];
+        severity?: "error" | "critical" | "info" | "warning" | undefined;
+    }, {
+        value: number;
+        type: "performance" | "accuracy" | "completeness" | "consistency" | "efficiency" | "reliability";
+        description: string;
+        signal_id: string;
+        severity?: "error" | "critical" | "info" | "warning" | undefined;
+        evidence?: string[] | undefined;
+    }>, "many">;
+    /** Extracted learning signals */
+    learning_signals: z.ZodArray<z.ZodObject<{
+        /** Unique identifier for this learning */
+        learning_id: z.ZodString;
+        /** Category of learning */
+        category: z.ZodEnum<["pattern", "anti_pattern", "optimization", "edge_case", "dependency", "constraint"]>;
+        /** Brief title for the learning */
+        title: z.ZodString;
+        /** Detailed description of the learning */
+        description: z.ZodString;
+        /** Confidence in this learning (0.0 - 1.0) */
+        confidence: z.ZodNumber;
+        /** Affected agents or components */
+        affected_agents: z.ZodDefault<z.ZodArray<z.ZodString, "many">>;
+        /** Recommendations for applying this learning (informational only) */
+        recommendations: z.ZodDefault<z.ZodArray<z.ZodString, "many">>;
+    }, "strip", z.ZodTypeAny, {
+        confidence: number;
+        description: string;
+        learning_id: string;
+        category: "pattern" | "anti_pattern" | "optimization" | "edge_case" | "dependency" | "constraint";
+        title: string;
+        affected_agents: string[];
+        recommendations: string[];
+    }, {
+        confidence: number;
+        description: string;
+        learning_id: string;
+        category: "pattern" | "anti_pattern" | "optimization" | "edge_case" | "dependency" | "constraint";
+        title: string;
+        affected_agents?: string[] | undefined;
+        recommendations?: string[] | undefined;
+    }>, "many">;
+    /** Identified gaps */
+    gap_analysis: z.ZodArray<z.ZodObject<{
+        /** Unique identifier for this gap */
+        gap_id: z.ZodString;
+        /** Type of gap */
+        type: z.ZodEnum<["coverage", "capability", "data", "process", "integration", "documentation"]>;
+        /** Brief title for the gap */
+        title: z.ZodString;
+        /** Detailed description of the gap */
+        description: z.ZodString;
+        /** Impact assessment */
+        impact: z.ZodEnum<["low", "medium", "high", "critical"]>;
+        /** Steps involved where gap was identified */
+        affected_steps: z.ZodDefault<z.ZodArray<z.ZodString, "many">>;
+        /** Evidence supporting gap identification */
+        evidence: z.ZodDefault<z.ZodArray<z.ZodString, "many">>;
+    }, "strip", z.ZodTypeAny, {
+        type: "data" | "coverage" | "capability" | "process" | "integration" | "documentation";
+        description: string;
+        affected_steps: string[];
+        evidence: string[];
+        title: string;
+        gap_id: string;
+        impact: "low" | "medium" | "high" | "critical";
+    }, {
+        type: "data" | "coverage" | "capability" | "process" | "integration" | "documentation";
+        description: string;
+        title: string;
+        gap_id: string;
+        impact: "low" | "medium" | "high" | "critical";
+        affected_steps?: string[] | undefined;
+        evidence?: string[] | undefined;
+    }>, "many">;
+    /** Cross-event correlations found */
+    correlations: z.ZodDefault<z.ZodArray<z.ZodObject<{
+        /** Correlation identifier */
+        correlation_id: z.ZodString;
+        /** Type of correlation */
+        type: z.ZodEnum<["causal", "temporal", "similarity", "dependency"]>;
+        /** Events involved */
+        event_refs: z.ZodArray<z.ZodString, "many">;
+        /** Description of the correlation */
+        description: z.ZodString;
+        /** Correlation strength (0.0 - 1.0) */
+        strength: z.ZodNumber;
+    }, "strip", z.ZodTypeAny, {
+        type: "dependency" | "causal" | "temporal" | "similarity";
+        description: string;
+        correlation_id: string;
+        event_refs: string[];
+        strength: number;
+    }, {
+        type: "dependency" | "causal" | "temporal" | "similarity";
+        description: string;
+        correlation_id: string;
+        event_refs: string[];
+        strength: number;
+    }>, "many">>;
+    /** Summary statistics */
+    summary: z.ZodObject<{
+        /** Overall quality score across all events (0.0 - 1.0) */
+        overall_quality_score: z.ZodNumber;
+        /** Total quality signals extracted */
+        total_quality_signals: z.ZodNumber;
+        /** Total learning signals extracted */
+        total_learning_signals: z.ZodNumber;
+        /** Total gaps identified */
+        total_gaps: z.ZodNumber;
+        /** Events meeting expectations percentage */
+        expectations_met_rate: z.ZodNumber;
+        /** Key findings (informational summaries) */
+        key_findings: z.ZodArray<z.ZodString, "many">;
+        /** Improvement suggestions (informational only - NOT actionable by this agent) */
+        improvement_suggestions: z.ZodArray<z.ZodString, "many">;
+    }, "strip", z.ZodTypeAny, {
+        overall_quality_score: number;
+        total_quality_signals: number;
+        total_learning_signals: number;
+        total_gaps: number;
+        expectations_met_rate: number;
+        key_findings: string[];
+        improvement_suggestions: string[];
+    }, {
+        overall_quality_score: number;
+        total_quality_signals: number;
+        total_learning_signals: number;
+        total_gaps: number;
+        expectations_met_rate: number;
+        key_findings: string[];
+        improvement_suggestions: string[];
+    }>;
+    /** Reflection metadata */
+    version: z.ZodDefault<z.ZodString>;
+}, "strip", z.ZodTypeAny, {
+    version: string;
+    summary: {
+        overall_quality_score: number;
+        total_quality_signals: number;
+        total_learning_signals: number;
+        total_gaps: number;
+        expectations_met_rate: number;
+        key_findings: string[];
+        improvement_suggestions: string[];
+    };
+    reflection_id: string;
+    events_analyzed: number;
+    agents_analyzed: string[];
+    analysis_time_range: {
+        earliest: string;
+        latest: string;
+    };
+    outcome_evaluations: {
+        agent_id: string;
+        decision_type: string;
+        decision_ref: string;
+        outcome_score: number;
+        summary: string;
+        dimensions: {
+            name: string;
+            score: number;
+            notes?: string | undefined;
+        }[];
+        met_expectations: boolean;
+        deviation_notes?: string | undefined;
+    }[];
+    quality_signals: {
+        value: number;
+        type: "performance" | "accuracy" | "completeness" | "consistency" | "efficiency" | "reliability";
+        description: string;
+        signal_id: string;
+        evidence: string[];
+        severity?: "error" | "critical" | "info" | "warning" | undefined;
+    }[];
+    learning_signals: {
+        confidence: number;
+        description: string;
+        learning_id: string;
+        category: "pattern" | "anti_pattern" | "optimization" | "edge_case" | "dependency" | "constraint";
+        title: string;
+        affected_agents: string[];
+        recommendations: string[];
+    }[];
+    gap_analysis: {
+        type: "data" | "coverage" | "capability" | "process" | "integration" | "documentation";
+        description: string;
+        affected_steps: string[];
+        evidence: string[];
+        title: string;
+        gap_id: string;
+        impact: "low" | "medium" | "high" | "critical";
+    }[];
+    correlations: {
+        type: "dependency" | "causal" | "temporal" | "similarity";
+        description: string;
+        correlation_id: string;
+        event_refs: string[];
+        strength: number;
+    }[];
+}, {
+    summary: {
+        overall_quality_score: number;
+        total_quality_signals: number;
+        total_learning_signals: number;
+        total_gaps: number;
+        expectations_met_rate: number;
+        key_findings: string[];
+        improvement_suggestions: string[];
+    };
+    reflection_id: string;
+    events_analyzed: number;
+    agents_analyzed: string[];
+    analysis_time_range: {
+        earliest: string;
+        latest: string;
+    };
+    outcome_evaluations: {
+        agent_id: string;
+        decision_type: string;
+        decision_ref: string;
+        outcome_score: number;
+        summary: string;
+        met_expectations: boolean;
+        dimensions?: {
+            name: string;
+            score: number;
+            notes?: string | undefined;
+        }[] | undefined;
+        deviation_notes?: string | undefined;
+    }[];
+    quality_signals: {
+        value: number;
+        type: "performance" | "accuracy" | "completeness" | "consistency" | "efficiency" | "reliability";
+        description: string;
+        signal_id: string;
+        severity?: "error" | "critical" | "info" | "warning" | undefined;
+        evidence?: string[] | undefined;
+    }[];
+    learning_signals: {
+        confidence: number;
+        description: string;
+        learning_id: string;
+        category: "pattern" | "anti_pattern" | "optimization" | "edge_case" | "dependency" | "constraint";
+        title: string;
+        affected_agents?: string[] | undefined;
+        recommendations?: string[] | undefined;
+    }[];
+    gap_analysis: {
+        type: "data" | "coverage" | "capability" | "process" | "integration" | "documentation";
+        description: string;
+        title: string;
+        gap_id: string;
+        impact: "low" | "medium" | "high" | "critical";
+        affected_steps?: string[] | undefined;
+        evidence?: string[] | undefined;
+    }[];
+    version?: string | undefined;
+    correlations?: {
+        type: "dependency" | "causal" | "temporal" | "similarity";
+        description: string;
+        correlation_id: string;
+        event_refs: string[];
+        strength: number;
+    }[] | undefined;
+}>;
+export type ReflectionOutput = z.infer<typeof ReflectionOutputSchema>;