Jesús Pérez d14150da75 feat: Phase 5.3 - Multi-Agent Learning Infrastructure
Implement intelligent agent learning from Knowledge Graph execution history
with per-task-type expertise tracking, recency bias, and learning curves.

## Phase 5.3 Implementation

### Learning Infrastructure ( Complete)
- LearningProfileService with per-task-type expertise metrics
- TaskTypeExpertise model tracking success_rate, confidence, learning curves
- Recency bias weighting: recent 7 days weighted 3x higher (exponential decay)
- Confidence scoring prevents overfitting: min(1.0, executions / 20)
- Learning curves computed from daily execution windows

### Agent Scoring Service ( Complete)
- Unified AgentScore combining SwarmCoordinator + learning profiles
- Scoring formula: 0.3*base + 0.5*expertise + 0.2*confidence
- Rank agents by combined score for intelligent assignment
- Support for recency-biased scoring (recent_success_rate)
- Methods: rank_agents, select_best, rank_agents_with_recency

### KG Integration ( Complete)
- KGPersistence::get_executions_for_task_type() - query by agent + task type
- KGPersistence::get_agent_executions() - all executions for agent
- Coordinator::load_learning_profile_from_kg() - core KG→Learning integration
- Coordinator::load_all_learning_profiles() - batch load for multiple agents
- Convert PersistedExecution → ExecutionData for learning calculations

### Agent Assignment Integration ( Complete)
- AgentCoordinator uses learning profiles for task assignment
- extract_task_type() infers task type from title/description
- assign_task() scores candidates using AgentScoringService
- Fallback to load-based selection if no learning data available
- Learning profiles stored in coordinator.learning_profiles RwLock

### Profile Adapter Enhancements ( Complete)
- create_learning_profile() - initialize empty profiles
- add_task_type_expertise() - set task-type expertise
- update_profile_with_learning() - update swarm profiles from learning

## Files Modified

### vapora-knowledge-graph/src/persistence.rs (+30 lines)
- get_executions_for_task_type(agent_id, task_type, limit)
- get_agent_executions(agent_id, limit)

### vapora-agents/src/coordinator.rs (+100 lines)
- load_learning_profile_from_kg() - core KG integration method
- load_all_learning_profiles() - batch loading for agents
- assign_task() already uses learning-based scoring via AgentScoringService

### Existing Complete Implementation
- vapora-knowledge-graph/src/learning.rs - calculation functions
- vapora-agents/src/learning_profile.rs - data structures and expertise
- vapora-agents/src/scoring.rs - unified scoring service
- vapora-agents/src/profile_adapter.rs - adapter methods

## Tests Passing
- learning_profile: 7 tests 
- scoring: 5 tests 
- profile_adapter: 6 tests 
- coordinator: learning-specific tests 

## Data Flow
1. Task arrives → AgentCoordinator::assign_task()
2. Extract task_type from description
3. Query KG for task-type executions (load_learning_profile_from_kg)
4. Calculate expertise with recency bias
5. Score candidates (SwarmCoordinator + learning)
6. Assign to top-scored agent
7. Execution result → KG → Update learning profiles

## Key Design Decisions
 Recency bias: 7-day half-life with 3x weight for recent performance
 Confidence scoring: min(1.0, total_executions / 20) prevents overfitting
 Hierarchical scoring: 30% base load, 50% expertise, 20% confidence
 KG query limit: 100 recent executions per task-type for performance
 Async loading: load_learning_profile_from_kg supports concurrent loads

## Next: Phase 5.4 - Cost Optimization
Ready to implement budget enforcement and cost-aware provider selection.
2026-01-11 13:03:53 +00:00

235 lines
6.2 KiB
Rust

// vapora-backend: Audit trail system
// Phase 3: Track all workflow events and actions
use chrono::{DateTime, Utc};
use serde::{Deserialize, Serialize};
use std::sync::Arc;
use tokio::sync::RwLock;
#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct AuditEntry {
pub id: String,
pub timestamp: DateTime<Utc>,
pub workflow_id: String,
pub event_type: String,
pub actor: String,
pub details: serde_json::Value,
}
impl AuditEntry {
pub fn new(
workflow_id: String,
event_type: String,
actor: String,
details: serde_json::Value,
) -> Self {
Self {
id: uuid::Uuid::new_v4().to_string(),
timestamp: Utc::now(),
workflow_id,
event_type,
actor,
details,
}
}
}
/// Audit trail maintains history of workflow events
pub struct AuditTrail {
entries: Arc<RwLock<Vec<AuditEntry>>>,
}
impl AuditTrail {
pub fn new() -> Self {
Self {
entries: Arc::new(RwLock::new(Vec::new())),
}
}
/// Log a workflow event
pub async fn log_event(
&self,
workflow_id: String,
event_type: String,
actor: String,
details: serde_json::Value,
) {
let entry = AuditEntry::new(workflow_id, event_type, actor, details);
let mut entries = self.entries.write().await;
entries.push(entry);
}
/// Get audit entries for a workflow
pub async fn get_workflow_audit(&self, workflow_id: &str) -> Vec<AuditEntry> {
let entries = self.entries.read().await;
entries
.iter()
.filter(|e| e.workflow_id == workflow_id)
.cloned()
.collect()
}
/// Get all audit entries
pub async fn get_all_entries(&self) -> Vec<AuditEntry> {
let entries = self.entries.read().await;
entries.clone()
}
/// Get entries by event type
pub async fn get_by_event_type(&self, event_type: &str) -> Vec<AuditEntry> {
let entries = self.entries.read().await;
entries
.iter()
.filter(|e| e.event_type == event_type)
.cloned()
.collect()
}
/// Get entries by actor
pub async fn get_by_actor(&self, actor: &str) -> Vec<AuditEntry> {
let entries = self.entries.read().await;
entries
.iter()
.filter(|e| e.actor == actor)
.cloned()
.collect()
}
/// Clear all entries (for testing)
pub async fn clear(&self) {
let mut entries = self.entries.write().await;
entries.clear();
}
}
impl Default for AuditTrail {
fn default() -> Self {
Self::new()
}
}
/// Event types for audit trail
pub mod events {
pub const WORKFLOW_CREATED: &str = "workflow_created";
pub const WORKFLOW_STARTED: &str = "workflow_started";
pub const WORKFLOW_COMPLETED: &str = "workflow_completed";
pub const WORKFLOW_FAILED: &str = "workflow_failed";
pub const WORKFLOW_ROLLED_BACK: &str = "workflow_rolled_back";
pub const PHASE_STARTED: &str = "phase_started";
pub const PHASE_COMPLETED: &str = "phase_completed";
pub const STEP_STARTED: &str = "step_started";
pub const STEP_COMPLETED: &str = "step_completed";
pub const STEP_FAILED: &str = "step_failed";
}
#[cfg(test)]
mod tests {
use super::*;
#[tokio::test]
async fn test_audit_trail_creation() {
let audit = AuditTrail::new();
assert!(audit.get_all_entries().await.is_empty());
}
#[tokio::test]
async fn test_log_event() {
let audit = AuditTrail::new();
audit
.log_event(
"wf-1".to_string(),
events::WORKFLOW_STARTED.to_string(),
"system".to_string(),
serde_json::json!({"test": "data"}),
)
.await;
let entries = audit.get_all_entries().await;
assert_eq!(entries.len(), 1);
assert_eq!(entries[0].workflow_id, "wf-1");
assert_eq!(entries[0].event_type, events::WORKFLOW_STARTED);
}
#[tokio::test]
async fn test_get_workflow_audit() {
let audit = AuditTrail::new();
audit
.log_event(
"wf-1".to_string(),
events::WORKFLOW_STARTED.to_string(),
"system".to_string(),
serde_json::json!({}),
)
.await;
audit
.log_event(
"wf-2".to_string(),
events::WORKFLOW_STARTED.to_string(),
"system".to_string(),
serde_json::json!({}),
)
.await;
let entries = audit.get_workflow_audit("wf-1").await;
assert_eq!(entries.len(), 1);
assert_eq!(entries[0].workflow_id, "wf-1");
}
#[tokio::test]
async fn test_filter_by_event_type() {
let audit = AuditTrail::new();
audit
.log_event(
"wf-1".to_string(),
events::WORKFLOW_STARTED.to_string(),
"system".to_string(),
serde_json::json!({}),
)
.await;
audit
.log_event(
"wf-1".to_string(),
events::WORKFLOW_COMPLETED.to_string(),
"system".to_string(),
serde_json::json!({}),
)
.await;
let entries = audit.get_by_event_type(events::WORKFLOW_STARTED).await;
assert_eq!(entries.len(), 1);
assert_eq!(entries[0].event_type, events::WORKFLOW_STARTED);
}
#[tokio::test]
async fn test_filter_by_actor() {
let audit = AuditTrail::new();
audit
.log_event(
"wf-1".to_string(),
events::WORKFLOW_STARTED.to_string(),
"user-1".to_string(),
serde_json::json!({}),
)
.await;
audit
.log_event(
"wf-2".to_string(),
events::WORKFLOW_STARTED.to_string(),
"user-2".to_string(),
serde_json::json!({}),
)
.await;
let entries = audit.get_by_actor("user-1").await;
assert_eq!(entries.len(), 1);
assert_eq!(entries[0].actor, "user-1");
}
}