116 lines
5.5 KiB
Rust
Raw Normal View History

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
// Agents marketplace page
use leptos::prelude::*;
use leptos::task::spawn_local;
use log::warn;
use crate::api::{Agent, ApiClient};
use crate::components::{Badge, Button, Card, GlowColor, NavBar};
use crate::config::AppConfig;
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
/// Agents marketplace page
#[component]
pub fn AgentsPage() -> impl IntoView {
let api_client = ApiClient::new(&AppConfig::load());
let (agents, set_agents) = signal(Vec::<Agent>::new());
let (loading, set_loading) = signal(true);
let (error, set_error) = signal(None::<String>);
// Fetch agents on mount
Effect::new(move |_| {
let api = api_client.clone();
spawn_local(async move {
match api.fetch_agents().await {
Ok(a) => {
set_agents.set(a);
set_loading.set(false);
}
Err(e) => {
warn!("Failed to fetch agents: {}", e);
set_error.set(Some(e));
set_loading.set(false);
}
}
});
});
view! {
<div class="min-h-screen bg-gradient-to-br from-slate-900 via-slate-800 to-slate-900">
<NavBar />
<div class="container mx-auto px-6 py-8">
<h1 class="text-3xl font-bold text-white mb-8">"Agent Marketplace"</h1>
<Show
when=move || !loading.get()
fallback=|| view! {
<div class="text-center py-12">
<div class="text-xl text-white">"Loading agents..."</div>
</div>
}
>
{move || {
if let Some(err) = error.get() {
view! {
<div class="bg-red-500/20 border border-red-500/50 rounded-lg p-6 text-center">
<div class="text-red-400 font-semibold mb-2">"Error loading agents"</div>
<div class="text-sm text-red-300">{err}</div>
</div>
}.into_any()
} else if agents.get().is_empty() {
view! {
<div class="text-center py-12">
<div class="text-xl text-gray-400">"No agents available"</div>
</div>
}.into_any()
} else {
view! {
<div class="grid grid-cols-1 md:grid-cols-2 lg:grid-cols-3 gap-6">
<For
each=move || agents.get()
key=|agent| agent.id.clone()
children=move |agent| {
let name = agent.name.clone();
let role = format!("{:?}", agent.role);
let llm_info = format!("LLM: {} {}", agent.llm_provider, agent.llm_model);
let capabilities = agent.capabilities.clone();
view! {
<Card glow=GlowColor::Cyan hover_effect=true>
<div class="flex items-start justify-between mb-3">
<h3 class="text-lg font-semibold text-white">
{name}
</h3>
<Badge class="bg-cyan-500/20 text-cyan-400 text-xs">
{role}
</Badge>
</div>
<p class="text-gray-400 text-sm mb-4">
{llm_info}
</p>
<div class="flex gap-2 flex-wrap mb-4">
{capabilities.iter().take(3).map(|cap| {
let capability = cap.clone();
view! {
<Badge class="bg-green-500/20 text-green-400 text-xs">
{capability}
</Badge>
}
}).collect_view()}
</div>
<Button class="w-full">
"View Agent"
</Button>
</Card>
}
}
/>
</div>
}.into_any()
}
}}
</Show>
</div>
</div>
}
}